Data Science Digest

2021-05-12 09:00:19 ​​Probabilistic Machine Learning Course by Philipp Hennig

The course by Philipp Hennig at the University of Tübingen covers the probabilistic paradigm for machine learning, and occasionally draws direct connections to statistical and deep learning. The course is aimed at master students in computer science and related fields.

https://bit.ly/3hkxuIU Open / Comment

2021-05-11 09:00:19 ​​What Is Face Recognition?

In this 101 tutorial, Adrian Rosebrock of the PyImageSearch team explains everything you need to know about face recognition, from what it is and how it works to how it is different from face detection and advanced face recognition algorithms you can start using today.

https://bit.ly/33y5qJQ Open / Comment

2021-05-10 09:00:10 ​​AutoNLP: Automatic Text Classification with SOTA Models

Developing NLP models can be challenging as you need to account for multiple factors, including model selection, data preprocessing, training, optimization, and infrastructure. AutoNLP, a tool to automate the end-to-end life cycle of an NLP model, can make this process much easier. Learn how to use AutoNLP in this step-by-step guide.

https://bit.ly/3xYkyhr Open / Comment

2021-05-09 09:30:09 ​​Multiple Time Series Forecasting with PyCaret

PyCaret is a popular machine learning library and a model management tool for automating machine learning workflows. It allows us to build and deploy end-to-end ML prototypes quickly and efficiently. In this step-by-step tutorial, you will learn how to use PyCaret to forecast multiple time series in less than 50 lines of code.

https://bit.ly/3xWy2KO Open / Comment

2021-05-08 12:43:35 ​​How to Plot XGBoost Trees in R

XGBoost is a popular ML algorithm, which is frequently used in Kaggle competitions and has many practical use cases. If you always wanted to learn more about XGBoost, this short tutorial is for you. You will learn how to prepare the dataset for modeling, train the XGBoot model, plot the XGBoot trees, then export tree plots, and plot multiple trees at once.

https://bit.ly/33pYiiv Open / Comment

2021-05-07 10:00:12 ​​Motion Representations for Articulated Animation

In this research, Aliaksandr Siarohin et al. present novel motion representations for animating articulated objects consisting of distinct parts. Learn about the new method they propose, how it differs from keypoint-based works, and how it can be used to animate a variety of objects, surpassing previous methods on existing benchmarks.

Paper — https://bit.ly/3eVsVlk
Code — https://bit.ly/33q5nj4
Video — https://bit.ly/3tmvlOZ Open / Comment

2021-05-06 10:30:07 ​​Improving Model Performance Through Human Participation

In this article, Preetam Josh (Netflix) and Mudit Jain (Google) explore a complex topic of AI-to-human cooperation. Specifically, they explain how human input in the model inference loop (human-in-the-loop) can increase the final precision and recall, and how to incorporate human feedback at inference time to ensure higher precision and recall.

https://bit.ly/3eUWm7b Open / Comment

2021-05-06 09:30:09 ​​Data Science Digest — 05.05.21

The new issue of DataScienceDigest is here! Hop to learn about the latest articles, tutorials, research papers, and projects on DataScience, AI, ML, and BigData. All sections are prioritized for your convenience. Enjoy!

https://bit.ly/33mYRd3

Join @DataScienceDigest Open / Comment

2021-05-05 10:00:09 ​​Face Detection Tips, Suggestions, and Best Practices

In this tutorial, Adrian Rosenbrock continues to explore the topic of face detection. You will learn their tips, suggestions, and best practices to achieve high face detection accuracy with OpenCV and dlib. Though the tutorial is mostly theoretical, it features code and tons of useful links inside.

https://bit.ly/3ehR0na Open / Comment

2021-05-04 10:00:19 ​​skweak: Weak Supervision Made Easy for NLP

In this paper, Pierre Lison et al. present skweak, a versatile, Python-based software toolkit to help NLP developers apply weak supervision to a wide range of NLP tasks. The toolkit makes it easy to implement a large spectrum of labeling functions (such as heuristics, gazetteers, neural models, or linguistic constraints) on text data, apply them on a corpus, and aggregate their results in a fully unsupervised fashion.

Paper — https://bit.ly/3tk0ORU
Code — https://bit.ly/33aEmAj Open / Comment