You are reading the article 11 Superb Data Science Videos Every Data Scientist Must Watch updated in December 2023 on the website Daihoichemgio.com. We hope that the information we have shared is helpful to you. If you find the content interesting and meaningful, please share it with your friends and continue to follow and support us for the latest updates. Suggested January 2024 11 Superb Data Science Videos Every Data Scientist Must Watch
Overview
Presenting 11 data science videos that will enhance and expand your current skillset
We have categorized these videos into three fields – Natural Language Processing (NLP), Generative Models, and Reinforcement Learning
Learn how the concepts in these videos work and build your own data science project!
IntroductionI love learning and understanding data science concepts through videos. I simply do not have the time to pour through books and pages of text to understand different ideas and topics. Instead, I get a much better overview of concepts via videos and then pick and choose the topics I want to learn more about.
The sheer quality and diversity of topics available on platforms like YouTube never ceases to amaze. I recently learned about the amazing XLNet framework for NLP from a video (which I have mentioned below for your consumption). This helped me grasp the concept so I could explore more about XLNet!
I strongly believe structure is very necessary when we’re learning any concept or topic. I follow that approach each time I write an article as well. That’s why I’ve categorized these videos into their respective domains, primarily Natural Language Processing (NLP), Generative Models and Reinforcement Learning.
So are you ready to dive in and explore the length and breadth of data science through these fascinating videos?
Without any further ado, here are 11 awesome Data Science Videos:
Natural Language Processing (NLP)
XLNet explained
How does Google Duplex work?
Google’s POEMPORTRAITS: Combining Art and AI
Generative Models
Dive into Variational Autoencoders!
Create Facial Animation from Audio
MuseNet Learned to Compose Mozart, Bon Jovi, and More
Reinforcement Learning
Teaching the Computer to drive
Learn how AlphaGo Zero works
Google DeepMind AI learns to walk
AI learns to play 2048
BONUS
Adobe develops AI to detect Photoshopped Faces
XLNet ExplainedXLNet is the hottest framework in NLP right now. You simply must be aware of what it is and how it works if you want to carve out a career in this field. I came across this video recently and wanted to share it with the community as soon as possible.
XLNet is the latest state-of-the-art NLP framework. It has outperformed Google’s BERT on 20 NLP tasks and achieved state-of-the-art results on 18 of them. That is very, very impressive.
Make sure you check out our article covering XLNet and it’s powerful ability here.
The below video provides a clear explanation of the original XLNet research paper. Note: You might need to know a few NLP concepts beforehand to truly grasp the inner workings of XLNet.
How does Google Duplex work?Remember when Sundar Pichai went on stage and sent the whole world into a frenzy when he unveiled Google Duplex in his keynote at Google I/O 2023? I remember listening in complete awe to the super-realistic calls that the AI made.
It took a bit of time for the data science and NLP community to come up with an explanation as to how Google Duplex actually works. It’s pretty powerful and has the potential to change how we interact with machines.
So the million dollar question – did Google Duplex pass the Turing Test!? You decide after watching this video:
Google’s POEMPORTRAITS: Combining Art and AII am an artist and the prospect of combining any art form with Artificial Intelligence is extremely enticing. In a world where there is so much fear around AI, such applications are more than welcome.
Google’s POEMPORTRAITS AI has been trained on nineteenth-century poetry using NLP techniques. You can contribute and donate a word to generate your own POEMPORTRAIT. Check out how this awesome concept works:
Generative Models Dive into Variational Autoencoders!Here’s one of our favorite reinforcement learning experts Xander Streenbrugge from his wonderful ArxivInsights channel.
Variational Autoencoders (VAEs) are powerful generative models with diverse applications. You can generate human faces or synthesize your own music or use VAEs for removing noise from images.
I like this video a lot. Xander begins with an introduction to basic autoencoders and then goes into VAEs and disentangled beta-VAEs. Quite technical, but explained beautifully and concisely, in typical Xander style.
Xander is coming back to DataHack Summit this year so you can hear from him and meet him in person!
Create Facial Animation from AudioI was immediately drawn to the video when I read the title. This is Generative Models at their best! You can not only generate facial animation from audio but also generate different emotions for the same audio. And the facial expressions look incredibly natural.
If you aren’t following Two Minute Papers, you’re missing out. They regularly churn out videos breaking down the latest developments in easy-to-understand fashion. It’s a gem of a channel.
MuseNet Learned to Compose Mozart, Bon Jovi, and MoreOpenAI’s MuseNet is a deep neural network that generates musical compositions with different instruments and combines different styles. It uses the same general-purpose unsupervised technology as GPT-2 and the results are amazing.
Never heard of GPT-2? It’s an NLP framework on par with XLNet. Check out how MustNet works here:
Reinforcement Learning Teaching a Computer to DriveSelf-driving cars have always fascinated me. The sheer scale of an autonomous vehicle’s project is staggering. There are so many components, both on the hardware side as well as the data science side, that need to align for this project to work.
This is a perfect video for beginners to learn about Genetic Programming and Reinforcement Learning and how they are used to create powerful applications. Simon’s personality kept me hooked until the very end.
And I am definitely trying the project on my own.
Learn how Google DeepMind’s AlphaGo Zero worksAnother great video by Xander. He explains Google DeepMind’s popular paper on AlphaGo Zero.
AlphaGo Zero is a new version of the original AlphaGo program that beat human champion Lee Sedol comprehensively. I recommend reading our article on Monte Carlo Tree Search, the algorithm behind AlphaGo before proceeding to learn about AlphaGo Zero.
AlphaGo Zero uses Reinforcement Learning to beat the world’s leading Go players without using data from human games.
“AlphaGo Zero surpassed the strength of AlphaGo Lee in three days by winning 100 games to 0, reached the level of AlphaGo Master in 21 days, and exceeded all the old versions in 40 days.”
Source: Wikipedia
Google DeepMind’s AI learns to WalkThis video is both hilarious and informative. Exactly the type of video I like when I’m learning new things! It was funny to watch the AI learn to walk. But at the same time, it left me marveling at the power of Reinforcement Learning.
The video discusses 3 papers to try and explain how the AI learned to walk and it is surprisingly simple to understand.
AI learns to play 2048Have you ever played the 2048 game? It is super addictive once you get the hang of it. I used to easily finish games earlier but not anymore. Being a data science enthusiast, I am going to train my computer to play it with the help of this awesome video.
This is another example of the use of Genetic Programming and Evolutionary Algorithms.
BONUS: Adobe develops AI to detect Photoshopped FacesAdobe is a market leader in image and video manipulation software. Other companies have tried, but not many have even gotten close to Adobe’s level.
Last month, Adobe announced its research efforts to detect manipulated images. High time someone did that! It will soon be impossible to tell real from fake given how quickly GANs have taken over the world.
Imagine Donald Trump challenging Kim Jong Un to a nuclear war and then claiming that it was a deepfake and shrugging off all responsibility! We need to avoid those situations turning into reality. This video shows how Adobe’s algorithm works and tried to combat fake images:
End NotesI love knowing about the latest research in data science, machine learning and AI. But I find it hard to read papers. It takes a lot of time and effort – something not every data science professional has. I am sure many of you struggle with the same. Consuming videos is the ideal way to get an overview of these concepts.
You can then pick and choose where your interests lie and try to spin up a project or blog post on it. Trust me, it’s a wonderful way to learn and ingrain new data science concepts.
Related
You're reading 11 Superb Data Science Videos Every Data Scientist Must Watch
Data Scientist Vs Data Analyst: Key Differences Explained
In the world of data-driven decisions, two prominent roles have emerged: data analysts and data scientists. These professionals play a crucial role in helping organizations harness the power of data, but their responsibilities and skill sets are quite different.
Data analysts focus on using data visualization and statistical analysis to understand data and identify patterns. They are usually required to have at least a bachelor’s degree in a relevant field like mathematics, statistics, computer science, or finance.
Broadly speaking, both professions involve extracting valuable insights from data; however, their approaches and skill sets do vary.
In this article, we will explore the differences between data scientists and data analysts and highlight the unique skills and responsibilities required for each role.
Let’s dive in.
While data scientists and data analysts both work with data, they have distinct roles and responsibilities.
Understanding the differences between these two roles is important for organizations seeking to build an effective data team. Also, it is crucial for those that would like a career in data to understand.
In this section, we will explore the key differences between data scientists and data analysts, including their educational backgrounds, technical skills, and the types of problems they are typically tasked with solving.
The table below gives a quick overview of the differences between the two roles:
Education/BackgroundData ScientistData AnalystDegreeBachelor’s degree in business, economics, statistics, or a related fieldBachelor’s degree in business, economics, statistics, or related fieldProgramming skillsProficient in languages such as Python, R, and SQLProficient in Excel, SQL, and basic scripting languagesMathematics skillsStrong mathematical skills, including linear algebra, calculus, and statisticsStrong statistical skills, including regression analysis and hypothesis testingWork experienceExperience with big data technologies, machine learning, and data visualizationExperience with statistical analysis, data modeling, and reporting
Joining a boot camp, using tutorials, or completing online courses or certificate programs may not cut it.
Data scientists should have a strong foundation in mathematics, statistics, and computer science, as well as hands-on experience with programming languages such as Python, R, and SQL.
Other data analyst skills include working with databases and having basic scripting language skills.
The job involves working on large data sets, developing predictive models, and extracting insights from data. Like data analysts, it also requires soft skills like communication and collaboration since you often need to work with different teams.
Data analysts: Very simply, a data analyst’s job involves analyzing and interpreting data to provide insights and recommendations to stakeholders.
You may be tasked with working with different data sources to identify trends and patterns that can inform business decisions.
Some specific responsibilities of data analysts can include:
Collecting, cleaning, and organizing data from various sources
Conducting statistical analysis to identify trends and patterns in data using software like Tableau
Creating reports and dashboards to visualize data and communicate insights to stakeholders
Identifying areas for process improvement and making data-driven recommendations to stakeholders
Developing and maintaining databases and data systems to support data analysis
Keeping up-to-date with the latest trends and developments in data analysis and visualization.
Now, things get a little more complex.
Data scientists: Being a data scientist involves analyzing complex data sets, developing predictive models, and extracting insights from data.
They work closely with stakeholders across different departments to provide insights and recommendations based on their data analysis.
Some specific responsibilities of data scientists include:
Conducting exploratory data analysis to identify patterns and trends in data
Developing predictive models using statistical and machine learning techniques
Building and testing machine learning models to improve predictive accuracy
Using problem-solving skills and business intelligence to come up with data-driven solutions to business problems
Communicating complex findings and recommendations to non-technical stakeholders
Collaborating with data engineers and software developers to build and deploy data-driven solutions
In the next two sections, we’ll take a look at the future job prospects and salary expectations for the two professions.
The job outlook for data scientists in 2023 is very promising as organizations across industries continue to collect and analyze increasing amounts of data.
According to the U.S. Bureau of Labor Statistics (BLS), employment of data scientists is projected to grow by 36% from 2023 to 2031, which is much faster than the average when compared to other occupations. Job opportunities in the field are driven by the increasing use of data and analytics to drive decision-making in organizations of all sizes.
According to Glassdoor, the national average salary for data scientists in the United States is around $103,000 per year. Many organizations also offer various additional forms of compensation for data scientists, such as bonuses, equity, and other benefits like medical insurance and paid time off.
Please note that compensation can vary widely depending on location, industry, and years of experience.
According to the BLS, employment of management analysts (which includes data analyst careers) is projected to grow by 11% from 2023 to 2030. Like data scientists, the job outlook for data analysts is very positive for the foreseeable future.
Compensation for data analysts may vary based on factors such as experience, industry, and location. Entry-level data analysts typically earn lower salaries, they can expect their pay to increase as their skills and expertise develop over time.
In terms of salary, the national average for data analyst positions in the United States is around $65,850 per year, according to Glassdoor.
The job prospects and compensation for both data scientists and data analysts are very promising, but how can you decide which career is right for you? We’re going to take a look at factors to consider in the next section.
Deciding which career path is right for you can feel daunting, but think of it as an exciting opportunity to explore this wonderful world of data!
The two fields may seem similar at first glance, and in a way, they are, but they require different skill sets and offer unique career paths.
With the right information and guidance, you can choose the path that is best suited for your skills, interests, and career goals.
In this section, we’ll provide some tips and insights to help you navigate this decision and choose the right path for you.
When considering a career in data science or data analysis, it’s important to think about your skills, interests, and career goals.
Here are some specific factors to consider:
Roles and responsibilities: Data scientists are often responsible for more strategic and complex initiatives, such as developing predictive models or creating machine learning algorithms. Data analyst roles focus more on day-to-day operations and providing insights to stakeholders.
Job outlook and salary: Both data scientists and data analysts have strong job prospects and competitive salaries, but the specific job outlook and salary can vary depending on the industry, location, and years of experience.
Ultimately, the right path for you will come down to your individual goals and aspirations.
Now one great thing about data skills is that they can be applied in most industries, let check them out.
The field of data science and data analytics is in high demand across a wide range of industries and company types.
Here are some examples of industries that both commonly employ data scientists and data analysts:
Finance and Banking: The finance and banking industry relies heavily on data analytics to identify trends, assess risk, and make informed business decisions. Business analysts are in high demand.
Healthcare: Healthcare organizations use data science and data analytics to improve patient outcomes, manage resources, and drive innovation in medical research.
E-commerce: E-commerce companies use data analytics to better understand their customer’s behavior, preferences, and purchasing habits in order to improve marketing and sales strategies.
Technology: Technology companies use data science and data analytics to develop new products and services, improve user experiences, come up with real-world solutions, and identify areas for innovation and growth.
There are employment opportunities across different company types, including startups, large corporations, consulting firms, and government agencies.
Understanding the diverse range of industries and company types that rely on data professionals is crucial for individuals looking to build successful careers in these fields.
It’s also important to note that both fields are evolving, and there are emerging trends that are worth considering.
In addition to industry types, consider emerging trends in data science and data analytics that are changing the landscape of the two fields.
Here are some current trends that are shaping the future of data science and data analytics:
Artificial intelligence and machine learning: AI and machine learning are increasingly being used in data science and data analytics to automate data processing, identify patterns, and make predictions. These technologies have the potential to revolutionize industries from healthcare to finance to marketing.
Cloud computing: Cloud computing has made it easier and more cost-effective to store, manage, and analyze large amounts of data. As cloud infrastructure and technology continue to improve, it’s expected that cloud-based data analytics and machine learning will become more widespread.
Data ethics and privacy: As more and more data is collected and analyzed, concerns about data ethics and privacy have come to the forefront. Data scientists and analysts are being called upon to ensure that data is being used ethically and responsibly and to implement measures to protect sensitive data.
Internet of things (IoT): The IoT refers to the network of interconnected devices and sensors that collect and share data. With the increasing adoption of IoT technology, there is a growing need for data scientists and analysts who can manage and analyze the vast amounts of data generated by these devices.
In the world of data, both data scientists and data analysts play important full-time roles in a business. While there are similarities between the two, they possess distinct differences in terms of responsibilities and required skills.
Data analysts primarily focus on working with structured data to solve tangible business problems using SQL, R, or Python programming languages, data visualization tools, and statistical analysis. They help organizations identify trends and derive insights from data.
On the other hand, data scientists are more involved in programming machines, optimizing systems, and creating frameworks and algorithms for collecting usable data. Their primary duties lie in collecting data and designing robust data-driven solutions.
While both job descriptions work within the realm of big data, identifying the right path depends on your interests, skills, and career goals. Whichever path you choose, both data scientists and data analysts are in-demand careers, making them an exciting and rewarding choices for those interested in working with data.
Top Big Data/Data Science Job Openings In Adobe To Watch Out For This Month
Land a career in Adobe with these top big data/data science jobs.
Many businesses encountered turbulence in 2023, yet big data/data science saw substantial demand and growth.
Data science
professionals are in high demand all across the world. These job opportunities will continue to grow after 2023, with over 1.5 lakh more positions being added. This is a natural reaction to data’s importance as a resource for businesses in the digital age. We’ve compiled a list of the top 10
Big Data
/Data Science job openings in Adobe to watch out for this month.
Big Data DeveloperLocation:
Bangalore
Requirements:
5+ years in the design and development of large-scale data-driven systems.
Work experience with one or more big data technologies such as Apache Spark.
Work experience with one or more NoSQL storage systems such as Aerospike, HBase, Cassandra.
Contribution to open source is desirable.
Great problem solving, coding (in Java/Scala, etc.), and system design skills.
Know more
here
.
Data ScientistLocation:
Noida, Uttar Pradesh
Responsibilities:
Perform exploratory data analysis quickly, generate and test working hypotheses, and discover new trends and relationships.
Reports and presentations can be used to communicate results and educate others.
Know more
here
.
Senior Data EngineerLocation:
Bengaluru, Karnataka
Responsibilities:
Develop distributed data processing pipelines using Apache Spark. Build and maintain pipelines as needed to power critical business metrics to measure the performance of pages on the website.
Responsible for crafting, developing sophisticated data applications/pipelines on large-scale data platforms using Apache Spark, Hadoop, Python/Scala.
Know more
here
.
Computer Scientist – PythonLocation:
Bengaluru, Karnataka
Responsibilities:
Developing Java backend services that would make use of and add value to Adobe’s own data platform.
Building the company’s tracking services in a cookie-less world.
Know more
here
.
Web & Data Science AnalystLocation:
Noida, Uttar Pradesh
Responsibilities:
Selecting features, building and optimizing classifiers using machine learning techniques.
Data mining using state-of-the-art methods.
Doing ad-hoc analysis and communicating results in a clear manner.
Crafting automated anomaly detection systems and constant tracking of its performance.
Know more
here
.
Computer ScientistLocation:
San Francisco
Responsibilities:
Build high-performance and resilient micro-services for event and data processing at scale.
Design new features and create functional specifications by working with product management and engineering team members.
Develop software solutions by understanding the company’s customer’s requirements, data flows, and integration architectures.
Know more
here
.
Data Scientist/Senior Product Analyst, ExperimentationLocation:
San Jose
Responsibilities:
You will work with data engineers to design and automate data pipelines to scale experimentation and user analytics.
In collaboration with a multi-functional team of product management, marketing, and engineering, you will tap into the underlying data, align on metrics/methodologies and generate insights to develop valuable, highly effective programs.
Know more
here
.
Web Analyst & Data ScienceLocation:
Bangalore
Responsibilities:
Responsible for providing Analytical Insights & Intelligence support aligned towards business or project or initiative.
Drive partnership with US Web Analytics team, Go-To-Market teams, eCommerce teams, chúng tôi Product Managers team, etc., and be the Subject Matter Expert for aligned areas.
Know more
here
.
Adobe Analytics – Big Data Software DeveloperLocation:
Bucharest
Responsibilities:
Transform the business requirements into feature specifications.
Contribute to the design and implementation of new features.
Design and implement new features, APIs, unit and integration test suites.
Be involved in all the product development and delivery stages, as part of a unified engineering team.
Data EngineerLocation:
San Jose
Responsibilities:
Design, develop & tune data products, applications, and integrations on large-scale data platforms (Hadoop, Snowflake, Alteryx, SSIS, Kafka Streaming, Hana, SQL server) with an emphasis on performance, reliability, and scalability, and most of all quality.
Analyze the business needs, profile large data sets and build custom data models and applications to drive the Adobe business decision making and customers experience.
These 7 Signs Show You Have Data Scientist Potential!
Overview
Let’s have a look at these 7 signs to know whether you are a potential data scientist or not
IntroductionData Science has not just become the “Sexiest Job of the 21st Century” but one of the most exciting job roles. You get to make an impact at the company-wide level with the latest technology and algorithms.
How does a person know that he needs to pursue Data Science? What are the qualities of a Data Scientist that one must look for? He may be interested in coding, building new tech, is patient at debugging. For example, a mathematics aspirant knows he wants to become a mathematician due to his interest in mathematical concepts and a knack for solving problems. So what are the signs that define a data scientist and one must look within themselves to know if they are a potential data scientist or not?
The role of a data scientist is really crucial to the whole organization and the economy as a whole. But the problem is – there is a shortage of “Skilled” data scientists globally. The AI and ML Blackbelt+ program aims to make you an industry-ready certified data science professional with 14+ courses, 39+ real-life projects, and 1:1 mentorship sessions so that you are never off-track.
If you are new to the world of data science, I suggest you go through the data science roadmap –
Table of Contents
Love Number Crunching
You are always up for solving Puzzles
Enjoy solving unstructured problems
You are curious – always asking “Why?”
You have a knack for problem-solving
Enjoy deep research
Love telling Stories – Great at presenting
1. Love Number Crunching and Solving PuzzlesIf you love crunching numbers and solving logical problems based on probability, statistics, puzzles then chances are that you have a natural tendency of becoming a data scientist or a business analyst. By “love”, I don’t mean calculating the bill split among your friends accurately, but a craze that reaches to next level.
For example, guessing the number of boys in India that are under 15 years. This is known as guess estimate and Data science interviewers love asking these questions. I have found the perfect article for you to understand problem-solving as data scientists –
2. Enjoy solving unstructured problemsUnstructured problems are everywhere around us. In an Edtech company, the management may be asking – how do we increase the revenue on our courses? Whereas in a social networking company, the question maybe – How do we increase the user retention on our app. Do you notice something wrong with these questions? Well, these are unstructured problems.
An important aspect of being a data scientist is the ability to form well-defined goals. A structured approach to the above question can be – To retain 20% more customers through the slide-down feature in the next 3 months. The goal should represent What? Why? How?
Next time, if you find yourself in the middle of a problem and you find yourself breaking down the problem into smaller goals, this might be a good starting sign of becoming a data scientist.
3. You are curious – always asking “Why?”Do you find yourself questioning other people’s assumptions? Are you not able to end your day without asking “Why this? Why that? Why this over that?” You may be a natural fit to become a data scientist.
For example, in the above example, you may want to question the management and ask – Why do we want to retain customers? Can we retain only the paying customers? why focus on retention rather than user acquisition?
Some of the best data scientists would stop anyone and ask for a rationale if they are not clear – Why did you ask this question? What was your thought process? Why do you assume so? are just a few examples of these questions!
Do you find yourself in the middle of these signs? You love problem-solving but don’t love mathematics by heart? A lot of these qualities can be gained with practice. The AI and ML Blackbelt+ certified Data Science program aims to take you from zero-to-hero with its 14+ courses, 39+ real-life projects, and 1:1 mentorship sessions with experts!
4. You have a knack for problem-solvingThe easiest sign that tells you are a potential data scientist is the addiction to solving problems. The fridge is not working? You try to identify the root cause of the problem. The sales of your company have decreased the quarter, you try to understand the root cause of the problem.
Data Scientist’s role is not just to apply machine learning algorithms to build an accurate model, it is to formulate a problem statement, form a hypothesis, data analysis, data model, and then finding the best results and communicating to the management. Each of these steps requires a knack for solving problems.
5. Enjoy Deep ResearchA Data scientist needs to sit on a single problem statement for a long duration of time, going back and forth to the stakeholders to understand their requirements, trying out different hypotheses, mining data, different modeling techniques until the results have been achieved.
Are you someone who finds themselves not giving up until the problem has been solved? Studying and Googling and researching on a single problem. When was the last time you spent hours and hours immersed in solving a problem? Can you do that again and again?
6. Love telling Stories – Great at presentingThere is always a person in the room who tells amazing stories. No one can resist listening to him and he impresses everyone with his storytelling skills. Are you that person?
A Data scientist needs to be great at storytelling. What is the use of all the hard work, if he is not able to influence his stakeholders? Communicating with data and presenting stories backed by data is one of the most important elements in the life of a data scientist.
7. Love ExperimentationsDo you find yourself experimenting at any time in the day? I don’t mean science experiments. A part of being a data scientist is to be curious.
What is the best mode of transportation to the office? If I find traffic at the X traffic signal, what is my estimated time of arrival? What if I travel halfway through the bus and then take the metro at the signal where I find traffic 90% of the time? These problems seem trivial but a large part of your life is going to be filled with experimentations.
If you love to experiment and play with curiosity in your daily life then being a data scientist is going to be a fun job for you!
End NotesBeing a data scientist is one of the most exciting roles and I love being part of this awesome community.
Were you able to relate to the points mentioned above? You are probably going to enjoy being a data scientist and do some amazing work in this field. The AI and ML Blackbelt+ program aims to nurture your talent as a data scientist and make you an industry-ready professional. You can check out the program here.
Which point were you able to relate to the most?
Related
40 Questions To Test A Data Scientist On Deep Learning
Introduction
Deep Learning has made many practical applications of machine learning possible. Deep Learning breaks down tasks in a way that makes all kinds of applications possible. This skilltest was conducted to test your knowledge of deep learning concepts.
A total of 853 people registered for this skill test. The test was designed to test the conceptual knowledge of deep learning. If you are one of those who missed out on this skill test, here are the questions and solutions. You missed on the real time test, but can read this article to find out how you could have answered correctly.
Here are the leaderboard ranking for all the participants.
Overall ScoresBelow are the distribution scores, they will help you evaluate your performance.
You can access the final scores here. More than 270 people participated in the skill test and the highest score obtained was 38. Here are a few statistics about the distribution.
Mean Score: 15.05
Median Score: 18
Mode Score: 0
Useful ResourcesA Complete Guide on Getting Started with Deep Learning in Python
The Evolution and Core Concepts of Deep Learning & Neural Networks
Practical Guide to implementing Neural Networks in Python (using Theano)
Fundamentals of Deep Learning – Starting with Artificial Neural Network
An Introduction to Implementing Neural Networks using TensorFlow
Fine-tuning a Keras model using Theano trained Neural Network & Introduction to Transfer Learning
6 Deep Learning Applications a beginner can build in minutes (using Python)
Questions & Answers1) The difference between deep learning and machine learning algorithms is that there is no need of feature engineering in machine learning algorithms, whereas, it is recommended to do feature engineering first and then apply deep learning.
A) TRUE
B) FALSE
Solution: (B)
Deep learning itself does feature engineering whereas machine learning requires manual feature engineering.
2) Which of the following is a representation learning algorithm?
A) Neural network
B) Random Forest
C) k-Nearest neighbor
D) None of the above
Solution: (A)
Neural network converts data in such a form that it would be better to solve the desired problem. This is called representation learning.
3) Which of the following option is correct for the below-mentioned techniques?
AdaGrad uses first order differentiation
L-BFGS uses second order differentiation
AdaGrad uses second order differentiation
L-BFGS uses first order differentiation
A) 1 and 2
B) 3 and 4
C) 1 and 4
D) 2 and 3
Solution: (A)
Option A is correct.
4) Increase in size of a convolutional kernel would necessarily increase the performance of a convolutional neural network.
A) TRUE
B) FALSE
Solution: (B)
Kernel size is a hyperparameter and therefore by changing it we can increase or decrease performance.
Question Context
Now you want to use this model on different dataset which has images of only Ford Mustangs (aka car) and the task is to locate the car in an image.
5) Which of the following categories would be suitable for this type of problem?
A) Fine tune only the last couple of layers and change the last layer (classification layer) to regression layer
B) Freeze all the layers except the last, re-train the last layer
C) Re-train the model for the new dataset
D) None of these
Solution: (A)
6) Suppose you have 5 convolutional kernel of size 7 x 7 with zero padding and stride 1 in the first layer of a convolutional neural network. You pass an input of dimension 224 x 224 x 3 through this layer. What are the dimensions of the data which the next layer will receive?
A) 217 x 217 x 3
B) 217 x 217 x 8
C) 218 x 218 x 5
D) 220 x 220 x 7
Solution: (C)
7) Suppose we have a neural network with ReLU activation function. Let’s say, we replace ReLu activations by linear activations.
Would this new neural network be able to approximate an XNOR function?
Note: The neural network was able to approximate XNOR function with activation function ReLu.
A) Yes
B) No
Solution: (B)
If ReLU activation is replaced by linear activation, the neural network loses its power to approximate non-linear function.
8) Suppose we have a 5-layer neural network which takes 3 hours to train on a GPU with 4GB VRAM. At test time, it takes 2 seconds for single data point.
Now we change the architecture such that we add dropout after 2nd and 4th layer with rates 0.2 and 0.3 respectively.
What would be the testing time for this new architecture?
A) Less than 2 secs
B) Exactly 2 secs
C) Greater than 2 secs
D) Can’t Say
Solution: (B)
The changes is architecture when we add dropout only changes in the training, and not at test time.
9) Which of the following options can be used to reduce overfitting in deep learning models?
Add more data
Use data augmentation
Use architecture that generalizes well
Add regularization
Reduce architectural complexity
A) 1, 2, 3
B) 1, 4, 5
C) 1, 3, 4, 5
D) All of these
Solution: (D)
All of the above techniques can be used to reduce overfitting.
10) Perplexity is a commonly used evaluation technique when applying deep learning for NLP tasks. Which of the following statement is correct?
A) Higher the perplexity the better
B) Lower the perplexity the better
Solution: (B)
11) Suppose an input to Max-Pooling layer is given above. The pooling size of neurons in the layer is (3, 3).
What would be the output of this Pooling layer?
A) 3
B) 5
C) 5.5
D) 7
Solution: (D)
Max pooling works as follows, it first takes the input using the pooling size we defined, and gives out the highest activated input.
12) Suppose there is a neural network with the below configuration.
If we remove the ReLU layers, we can still use this neural network to model non-linear functions.
A) TRUE
B) FALSE
Solution: (B)
13) Deep learning can be applied to which of the following NLP tasks?
A) Machine translation
B) Sentiment analysis
C) Question Answering system
D) All of the above
Solution: (D)
Deep learning can be applied to all of the above-mentioned NLP tasks.
14) Scenario 1: You are given data of the map of Arcadia city, with aerial photographs of the city and its outskirts. The task is to segment the areas into industrial land, farmland and natural landmarks like river, mountains, etc.
Deep learning can be applied to Scenario 1 but not Scenario 2.
A) TRUE
B) FALSE
Solution: (B)
Scenario 1 is on Euclidean data and scenario 2 is on Graphical data. Deep learning can be applied to both types of data.
15) Which of the following is a data augmentation technique used in image recognition tasks?
Horizontal flipping
Random cropping
Random scaling
Color jittering
Random translation
Random shearing
A) 1, 2, 4
B) 2, 3, 4, 5, 6
C) 1, 3, 5, 6
D) All of these
Solution: (D)
16) Given an n-character word, we want to predict which character would be the n+1th character in the sequence. For example, our input is “predictio” (which is a 9 character word) and we have to predict what would be the 10th character.
Which neural network architecture would be suitable to complete this task?
A) Fully-Connected Neural Network
B) Convolutional Neural Network
C) Recurrent Neural Network
D) Restricted Boltzmann Machine
Solution: (C)
Recurrent neural network works best for sequential data. Therefore, it would be best for the task.
17) What is generally the sequence followed when building a neural network architecture for semantic segmentation for image?
A) Convolutional network on input and deconvolutional network on output
B) Deconvolutional network on input and convolutional network on output
Solution: (A)
18) Sigmoid was the most commonly used activation function in neural network, until an issue was identified. The issue is that when the gradients are too large in positive or negative direction, the resulting gradients coming out of the activation function get squashed. This is called saturation of the neuron.
That is why ReLU function was proposed, which kept the gradients same as before in the positive direction.
A ReLU unit in neural network never gets saturated.
A) TRUE
B) FALSE
Solution: (B)
ReLU can get saturated too. This can be on the negative side of x-axis.
19) What is the relationship between dropout rate and regularization?
Note: we have defined dropout rate as the probability of keeping a neuron active?
A) Higher the dropout rate, higher is the regularization
B) Higher the dropout rate, lower is the regularization
Solution: (B)
Higher dropout rate says that more neurons are active. So there would be less regularization.
20) What is the technical difference between vanilla backpropagation algorithm and backpropagation through time (BPTT) algorithm?
A) Unlike backprop, in BPTT we sum up gradients for corresponding weight for each time step
B) Unlike backprop, in BPTT we subtract gradients for corresponding weight for each time step
Solution: (A)
BPTT is used in context of recurrent neural networks. It works by summing up gradients for each time step
21) Exploding gradient problem is an issue in training deep networks where the gradient getS so large that the loss goes to an infinitely high value and then explodes.
What is the probable approach when dealing with “Exploding Gradient” problem in RNNs?
A) Use modified architectures like LSTM and GRUs
B) Gradient clipping
C) Dropout
D) None of these
Solution: (B)
To deal with exploding gradient problem, it’s best to threshold the gradient values at a specific point. This is called gradient clipping.
22) There are many types of gradient descent algorithms. Two of the most notable ones are l-BFGS and SGD. l-BFGS is a second order gradient descent technique whereas SGD is a first order gradient descent technique.
In which of the following scenarios would you prefer l-BFGS over SGD?
Data is sparse
Number of parameters of neural network are small
A) Both 1 and 2
B) Only 1
C) Only 2
D) None of these
Solution: (A)
l-BFGS works best for both of the scenarios.
23) Which of the following is not a direct prediction technique for NLP tasks?
A) Recurrent Neural Network
B) Skip-gram model
C) PCA
D) Convolutional neural network
Solution: (C)
24) Which of the following would be the best for a non-continuous objective during optimization in deep neural net?
A) L-BFGS
B) SGD
C) AdaGrad
D) Subgradient method
Solution: (D)
Other optimization algorithms might fail on non-continuous objectives, but sub-gradient method would not.
25) Which of the following is correct?
Dropout randomly masks the input weights to a neuron
Dropconnect randomly masks both input and output weights to a neuron
A) 1 is True and 2 is False
B) 1 is False and 2 is True
C) Both 1 and 2 are True
D) Both 1 and 2 are False
Solution: (D)
In dropout, neurons are dropped; whereas in dropconnect; connections are dropped. So both input and output weights will be rendered in useless, i.e. both will be dropped for a neuron. Whereas in dropconnect, only one of them should be dropped
26) While training a neural network for image recognition task, we plot the graph of training error and validation error for debugging.
What is the best place in the graph for early stopping?
A) A
B) B
C) C
D) D
Solution: (C)
You would “early stop” where the model is most generalized. Therefore option C is correct.
27) Research is going on to solve image inpainting problems using computer vision with deep learning. For this, which loss function would be appropriate for computing the pixel-wise region to be inpainted?
Image inpainting is one of those problems which requires human expertise for solving it. It is particularly useful to repair damaged photos or videos. Below is an example of input and output of an image inpainting example.
A) Euclidean loss
B) Negative-log Likelihood loss
C) Any of the above
Solution: (C)
Both A and B can be used as a loss function for image inpainting problem.
A) Sum of squared error with respect to inputs
B) Sum of squared error with respect to weights
C) Sum of squared error with respect to outputs
D) None of the above
Solution: (C)
29) Mini-Batch sizes when defining a neural network are preferred to be multiple of 2’s such as 256 or 512. What is the reason behind it?
A) Gradient descent optimizes best when you use an even number
B) Parallelization of neural network is best when the memory is used optimally
C) Losses are erratic when you don’t use an even number
D) None of these
Solution: (B)
30) Xavier initialization is most commonly used to initialize the weights of a neural network. Below is given the formula for initialization.
If weights at the start are small, then signals reaching the end will be too tiny.
If weights at the start are too large, signals reaching the end will be too large.
Weights from Xavier’s init are drawn from the Gaussian distribution.
Xavier’s init helps reduce vanishing gradient problem.
Xavier’s init is used to help the input signals reach deep into the network. Which of the following statements are true?
A) 1, 2, 4
B) 2, 3, 4
C) 1, 3, 4
D) 1, 2, 3
E) 1, 2, 3, 4
Solution: (D)
All of the above statements are true.
31) As the length of sentence increases, it becomes harder for a neural translation machine to perform as sentence meaning is represented by a fixed dimensional vector. To solve this, which of the following could we do?
A) Use recursive units instead of recurrent
B)Use attention mechanism
C) Use character level translation
D) None of these
Solution: (B)
32) A recurrent neural network can be unfolded into a full-connected neural network with infinite length.
A) TRUE
B) FALSE
Solution: (A)
Recurrent neuron can be thought of as a neuron sequence of infinite length of time steps.
33) Which of the following is a bottleneck for deep learning algorithm?
A) Data related to the problem
B) CPU to GPU communication
C) GPU memory
D) All of the above
Solution: (D)
Along with having the knowledge of how to apply deep learning algorithms, you should also know the implementation details. Therefore you should know that all the above mentioned problems are a bottleneck for deep learning algorithm.
34) Dropout is a regularization technique used especially in the context of deep learning. It works as following, in one iteration we first randomly choose neurons in the layers and masks them. Then this network is trained and optimized in the same iteration. In the next iteration, another set of randomly chosen neurons are selected and masked and the training continues.
A) Affine layer
B) Convolutional layer
C) RNN layer
D) None of these
Solution: (C)
Dropout does not work well with recurrent layer. You would have to modify dropout technique a bit to get good results.
35) Suppose your task is to predict the next few notes of song when you are given the preceding segment of the song.
For example:
The input given to you is an image depicting the music symbols as given below,
Your required output is an image of succeeding symbols.
Which architecture of neural network would be better suited to solve the problem?
A) End-to-End fully connected neural network
B) Convolutional neural network followed by recurrent units
C) Neural Turing Machine
D) None of these
Solution: (B)
CNN work best on image recognition problems, whereas RNN works best on sequence prediction. Here you would have to use best of both worlds!
36) When deriving a memory cell in memory networks, we choose to read values as vector values instead of scalars. Which type of addressing would this entail?
A) Content-based addressing
B) Location-based addressing
Solution: (A)
A) Affine layer
B) Strided convolutional layer
C) Fractional strided convolutional layer
D) ReLU layer
Solution: (C)
Option C is correct. Go through this link.
Question Context 38-40
GRU is a special type of Recurrent Neural Networks proposed to overcome the difficulties of classical RNNs. This is the paper in which they were proposed: “On the Properties of Neural Machine Translation: Encoder–Decoder Approaches. Read the full paper here.
38) Which of the following statements is true with respect to GRU?
Units with short-term dependencies have reset gate very active.
Units with long-term dependencies have update gate very active
A) Only 1
B) Only 2
C) None of them
D) Both 1 and 2
Solution: (D)
39) If calculation of reset gate in GRU unit is close to 0, which of the following would occur?
A) Previous hidden state would be ignored
B) Previous hidden state would be not be ignored
Solution: (A)
40) If calculation of update gate in GRU unit is close to 1, which of the following would occur?
A) Forgets the information for future time steps
B) Copies the information through many time steps
Solution: (B)
End Notes
If you missed out on this competition, make sure you complete in the ones coming up shortly. We are giving cash prizes worth $10,000+ during the month of April 2023.
If you have any questions or doubts feel free to post them below.
Check out all the upcoming skilltests here.Related
25+ Websites To Find Datasets For Data Science Projects
Introduction
If there is one sentence, which summarizes the essence of learning data science, it is this:
The best way to learn data science is to apply data science.
If you are a beginner, you improve tremendously with each new project you undertake. If you are an experienced data science professional, you already know what I am talking about.
If you think that the situation above applies to you – Don’t worry! you are just at the right place. This article will provide you a list of websites / resources from which you can use data to do your own (pet) projects or even create your own products.
How can you use these sources?There is no end to how you can use these data sources. The application and usage is only limited by your creativity and application.
The simplest way to use them is to create data stories and publishing them over web. This would not only improve your data and visualization skills, but also improve your structured thinking.
On the other hand, if you are thinking / working on a data based product, these datasets could add power to your product by providing additional / new input data.
So, go ahead, work on these projects and share them with the larger world to showcase your data prowess!
I have divided these sources in various sections to help you categorize data sources based on application. We start with simple, generic and easy to handle datasets and then move to huge / industry relevant datasets. We then provide links to dataset for specific purpose – Text Mining, Image classification, Recommendation engine etc. This should provide you a holistic list of data resources.
Simple & Generic datasets to get you started
chúng tôi – This is the home of the U.S. Government’s open data. The site contains more than 190,000 data points at time of publishing. These datasets vary from data about climate, education, energy, Finance and many more areas.
chúng tôi – This is the home of the Indian Government’s open data. Find data by various industries, climate, health care etc. You can check out a few visualizations for inspiration here. Depending on your country of residence, you can also follow similar websites from a few other websites – check them out.
World Bank – The open data from the World bank. The platform provides several tools like Open Data Catalog, world development indices, education indices etc.
RBI – Data available from the Reserve Bank of India. This includes several metrics on money market operations, balance of payments, use of banking and several products. A must go to site, if you come from BFSI domain in India.
Five Thirty Eight Datasets – Here is a link to datasets used by Five Thirty Eight in their stories. Each dataset includes the data, a dictionary explaining the data and the link to the story carried out by Five Thirty Eight. If you want to learn how to create data stories, it can’t get better than this.
Huge Datasets – things are getting serious now!
Amazon Web Services (AWS) datasets – Amazon provides a few big datasets, which can be used on their platform or on your local computers. You can also analyze the data in the cloud using EC2 and Hadoop via EMR. Popular datasets on Amazon include full Enron email dataset, Google Books n-grams, NASA NEX datasets, Million Songs dataset and many more. More information can be found here.
A few months back, Google Research Group released YouTube labeled dataset, which consists of 8 million YouTube video IDs and associated labels from 4800 visual entities. It comes with pre-computed, state-of-the-art vision features from billions of frames.
Datasets for predictive modeling & machine learning:
UCI Machine Learning Repository – UCI Machine Learning Repository is clearly the most famous data repository. It is usually the first place to go, if you are looking for datasets related to machine learning repositories. The datasets include a diverse range of datasets from popular datasets like Iris and Titanic survival to recent contributions like that of Air Quality and GPS trajectories. The repository contains more than 350 datasets with labels like domain, purpose of the problem (Classification / Regression). You can use these filters to identify good datasets for your need.
Kaggle Kaggle has come up with a platform, where people can donate datasets and other community members can vote and run Kernel / scripts on them. They have more than 350 datasets in total – with more than 200 as Featured datasets. While some of the initial datasets were usually present at other places, I have seen a few interesting datasets on the platform, not present at other places. Along with new datasets, another benefit of the interface is that you can see scripts and questions from community members on the same interface.
Analytics Vidhya You can participate and download datasets from our practice problems and hackathon problems. The problem datasets are based on real-life industry problems and are relatively smaller as they are meant for 2 – 7 days hackathons. While practice problems are available to people always, the hackathon problems become unavailable after the hackathons. So, you need to participate on the hackathon to get access to the datasets.
Quandl Quandl provides financial, economic and alternative data from various sources through their website / API or direct integration with a few tools. Their datasets are classified as Open or Premium. You can access all the open datasets for Free, but you need to pay for the premium datasets. If you search, you still get good datasets on the platform. Eg. Stock Exchange data from India is available for free.
Past KDD Cups KDD Cup is the annual Data Mining and Knowledge Discovery competition organized by ACM Special Interest Group on Knowledge Discovery and Data Mining. Archives includes datasets and instructions. Winners are available for most years.
Driven Data Driven Data finds real-world challenges where data science can be used to create a positive social impact. They then run online modeling competitions for data scientists to develop the best models to solve them. If you are interested in use of data science for social good – this is the place to be.
Image classification datasets
The MNIST Database – The most popular dataset for image recognition using hand-written digits. It includes 60,000 train examples and a test set of 10,000 examples. This serves as typically the first dataset to practice image recognition.
Chars74K – Here is the next level of evolution, if you have passed hand written digits. This dataset includes character recognition in natural images. The dataset contains 74,000 images and hence the name of the dataset.
Frontal Face Images If you have worked on previous 2 projects and are able to identify digits and characters, here is the next level of challenge in Image recognition – Frontal Face images. The images were collected by CMU & MIT and are arranged in four folders.
ImageNet Time to build something generic now. Image database organised according to the WordNet hierarchy (currently only the nouns). Each node of the hierarchy is depicted by hundreds of images. Currently, the collection has an average of over five hundred images per node (and increasing).
Text Classification datasets
Twitter Sentiment Analysis The Twitter Sentiment Analysis Dataset contains 1,578,627 classified tweets, each row is marked as 1 for positive sentiment and 0 for negative sentiment. The data is in turn based on a Kaggle competition and analysis by Nick Sanders.
Datasets for Recommendation Engine
MovieLens MovieLens is a web site that helps people find movies to watch. It has hundreds of thousands of registered users. They conduct online field experiments in MovieLens in the areas of automated content recommendation, recommendation interfaces, tagging-based recommenders. These datasets are available for download and can be used to create your own recommender systems.
Jester Datasets about online joke recommender system
Websites which Curate list of datasets from various sources:
KDNuggets – The dataset page on KDNuggets has long been a reference point for people looking for datasets out there. A really comprehensive list, however some of the sources no longer provide the datasets. So, you will need to apply your own prudence on the datasets and the sources.
Awesome Public Datasets A GitHub repository with a comprehensive list of datasets categorized by domain. Datasets are classified neatly in various domains, which is very helpful. However, there is no description about the datasets on the repository itself – which could have made it very useful.
Reddit Datasets Subreddit Since this is a community driven forum, it might come across a bit messy (compared to previous 2 sources). However, you can sort datasets by popularity / votes to see the most popular ones. Also, it has some interesting datasets and discussions.
End NotesLooking forward to hearing from you.
If you like what you just read & want to continue your analytics learning, subscribe to our emails, follow us on twitter or like our facebook page.Related
Update the detailed information about 11 Superb Data Science Videos Every Data Scientist Must Watch on the Daihoichemgio.com website. We hope the article's content will meet your needs, and we will regularly update the information to provide you with the fastest and most accurate information. Have a great day!