Daily Dose of Data Science – Day 1 – Keep yourself updated with the state of the art with PapersWithCode

Daily dose of data science is an initiative started to bring everyday tips, suggestions and recommendations for all enthusiasts who are practicing the art and science of Data Science! In this post, we will discuss more on the wonderful platform of paperswithcode that will get help you discover some of the fascinating research work done in the field of data science and artificial intelligence!

Welcome everyone to day 1 of the series, daily dose of data science! As you all might have understood from the sub title of this post, daily dose of data science is targeted for any level of data scientists, machine learning and artificial intelligence experts and enthusiasts with the objective of providing tips, recommendations and new ideas related to the field of data science. One of the key areas which distinguishes good data scientists and better data scientists, is the area of research in data science and machine learning. The field has shown a significant growth and breakthrough in the last 5 years, mostly because of the availability of good datasets, open forums, communities and platforms to support and share knowledge about latest and state of the art methods and its application for variety of business purposes. Thus, if you are a budding or an established data scientist, it is even more important nowadays to keep yourself updated with the latest and the ground-breaking research work, with additional resources to implement the latest ideas yourself and practicing new ideas or modifications on the open datasets which are available. All of these can be found at one stop forum called paperwithcode.com.

Paperswithcode was introduced with a mission to accelerate the scientific progress, especially in the fields of data science and artificial intelligence, by making research work easier to discover, reproduce and build on the existing state of the art. The pace at which the fields of machine learning and data science are progressing, it becomes very hard to keep up with the progress. Although the entire internet is overloaded with information around various aspects of these fields, yet it was hard to easily discover, reproduce and find the underlying benchmarking datasets used for some of the fascinating research works. So, the Papers with Code team, in collaboration with Facebook AI Research, provides an extremely helpful avenue to reduce the frictions for discovering, reproducing and extending research. Hence, this is one of my favourite places to go for whenever I am working on any type of problem which can be solved by AI and data science.

Now, what are some of key values we see from Papers with Code?

1. Identify some of the trending research work in the world of Data Science!

For data science and research leaders and for any general enthusiasts, keeping a track of the trending research work is a common goal. Papers with Code makes it very easy to discover the trending research publications, with the quick summary, detailed paper with the associated code and datasets right from its home page!

2. Find the state of the art work for each domain and for a specific type of task!

When you are working on a specific problem, it is very important for you to know the latest work for that specific task without getting overloaded in the ocean of information which is available online. Papers with Code makes it extremely easy to discover and browse through some of the latest work with benchmarking datasets and detailed paper and code by indexing by the domain and the associated tasks.

3. Identify benchmarking datasets for reproducing results of existing work and comparing for any novel work

Discovering and downloading benchmarking datasets to reproduce result of existing work and using it to compare any novel approach can be a very challenging work. But Papers with Code makes it seamless to search and download the popular benchmarking datasets and even give certain insights about the usage of the benchmarking datasets in various research work.

4. Strengthen your conceptual understanding about various methods and techniques

The methods section helps you to identify and search papers and research work based on the core concepts of techniques and methods used in the research work. For example, if I want to learn about Attention networks, Transformer architectures, or Distributed Reinforcement learning, I can easily browse through the Methods section, get the core definition of these concepts and then read some of the papers with the core concept and around the related concepts and improve my intuitive and conceptual understanding.

5. Fast track the discovery process with the seamless search option!

Last but not the least, the search option on the top gives an ability to search any topic, dataset, task, domain and methods used based on their keywords. Although this is a basic expectation for any platform, but it makes it easy to navigate across the ocean of information effectively!

Well these are some of the benefits that I have found while using paperwithcode.com and I strongly recommend all of you to start using the same forum for keeping yourself updated with the latest work around the field of Data Science and Artificial Intelligence! That’s all for today! Visit again for getting another daily dose of data science and please feel free to like, share, comment and subscribe to my posts if you find it helpful!