Rami Krispin

𝐍𝐞𝐰 𝐫𝐞𝐥𝐞𝐚𝐬𝐞 𝐟𝐨𝐫 𝐔𝐛𝐞𝐫'𝐬 𝐟𝐨𝐫𝐞𝐜𝐚𝐬𝐭𝐢𝐧𝐠 𝐥𝐢𝐛𝐫𝐚𝐫𝐲 📊🌈👇🏼

𝐎𝐫𝐛𝐢𝐭 is an open-source Python library for 𝐁𝐚𝐲𝐞𝐬𝐢𝐚𝐧 time series forecasting and inference applications developed by Uber Engineering. The library uses under the hood probabilistic programming languages with libraries such as PyStan, Pyro, and PyTorch to build the forecast estimators. The new release, version 1.1 includes the following new features and changes:
✨ 𝐊𝐞𝐫𝐧𝐞𝐥 𝐓𝐢𝐦𝐞-𝐛𝐚𝐬𝐞𝐝 𝐑𝐞𝐠𝐫𝐞𝐬𝐬𝐢𝐨𝐧 (𝐊𝐓𝐑) - new forecasting model was added to the library based on a KTR model. KTR model uses latent variables to define a smooth, time-varying representation of regression coefficients. Four tutorials about using KTR were added to the library documentation.
✨ 𝐖𝐨𝐫𝐤𝐟𝐥𝐨𝐰 - Improve the object-oriented workflow by changing the library's classes, defining three types:
Forecaster
Model Template
Estimator
✨ 𝐒𝐲𝐧𝐭𝐚𝐱 - As a result of the changes in the class structure, some changes made for some of the library functions syntax
✨ 𝐃𝐚𝐭𝐚𝐯𝐢𝐳 - Updating the model diagnostics tools adding supports for 𝐀𝐫𝐕𝐢𝐳 library data visualization functions such as density, pair, and trace plots

𝐋𝐢𝐜𝐞𝐧𝐬𝐞: Apache 2.0

𝐑𝐞𝐬𝐨𝐮𝐫𝐜𝐞𝐬 👇🏼👇🏼👇🏼
Documentation: https://lnkd.in/gj_Ab6_q
Source code: https://lnkd.in/gFWwhFB8
Release blog: https://lnkd.in/g4E8m-EU
Colab notebook: https://lnkd.in/gcZDVcMi

#python #timeseries #forecasting #bayesianstatistics #bayesian #statistics #machinelearning #ml #pytorch #datavisualization #dataviz

Natural Language Processing Course 🚀

If you are looking for an NLP course, The University of Texas at Austin offers its master-level NLP course online for free. The course CS388, by Prof Greg Durrett, focuses on the foundations of NLP, and it covers topics such as:
✅ Intro and Linear Classification
✅ Multiclass and Neural Classification
✅ Word Embeddings
✅ Language Modeling and Self-Attention
✅ Transformers and Decoding
✅ Modern Large Language Models
✅ Machine Translation, Summarization

Resources 📚
Video lectures 📽️: https://lnkd.in/gb7WZv5u
Course website 🔗: https://lnkd.in/gbxG_DHH

#nlp #machinelearning #deeplearning #llm #datascience

𝐍𝐞𝐰 𝐥𝐢𝐛𝐫𝐚𝐫𝐲 𝐟𝐨𝐫 𝐮𝐩𝐥𝐢𝐟𝐭 𝐦𝐨𝐝𝐞𝐥𝐢𝐧𝐠! 🎊📈🚀

Booking.com released 𝐔𝐩𝐥𝐢𝐟𝐭𝐌𝐋, a new #python library for scalable unconstrained and constrained uplift modeling from experimental data. The library leverages 𝐏𝐲𝐒𝐩𝐚𝐫𝐤 and 𝐇2𝐎 models as the framework for uplift models.

𝐖𝐡𝐚𝐭 𝐢𝐬 𝐮𝐩𝐥𝐢𝐟𝐭 𝐦𝐨𝐝𝐞𝐥𝐢𝐧𝐠?
Uplift modeling is a family of techniques for estimating the Conditional Average Treatment Effect (CATE) from experimental or observational data using machine learning. In particular, we are interested in estimating the causal effect of treatment T on the outcome Y of an individual characterized by features X. In experimental data with binary treatments and binary outcomes, this is equivalent to estimating Pr(Y=1 | T=1, X=x) - Pr(Y=1 | T=0, X=x).

Thanks to the library contributors - Javier Albert and Irene Teinemaa, and to Noa Barbiro and Dima Goldenberg for sharing!🙏

𝐋𝐢𝐜𝐞𝐧𝐬𝐞: Apache 2
𝐃𝐨𝐜𝐮𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧: https://lnkd.in/gPYUm3DZ
𝐒𝐨𝐮𝐫𝐜𝐞 𝐜𝐨𝐝𝐞: https://lnkd.in/ghGTMYZV
𝐄𝐱𝐚𝐦𝐩𝐥𝐞𝐬: https://lnkd.in/gdsQDyxB
𝐁𝐨𝐨𝐤𝐢𝐧𝐠.𝐜𝐨𝐦 𝐝𝐚𝐭𝐚 𝐬𝐜𝐢𝐞𝐧𝐜𝐞 𝐛𝐥𝐨𝐠: https://booking.ai/

#machinelearning #causalinference #experimentation #ml #datascience #datascientist #h2o #pyspark

𝐏𝐲𝐌𝐂 4.0 𝐢𝐬 𝐎𝐮𝐭! 🎉🎉🎉

𝐏𝐲𝐌𝐂 made a major update yesterday and released version 4.0 🌈.

PyMC is one of the main 𝐏𝐲𝐭𝐡𝐨𝐧 packages for 𝐁𝐚𝐲𝐞𝐬𝐢𝐚𝐧 modeling ❤️. It provides a framework for probabilistic programming enabling users to build Bayesian models with a simple Python API and fit them using 𝐌𝐚𝐫𝐤𝐨𝐯 𝐂𝐡𝐚𝐢𝐧 𝐌𝐨𝐧𝐭𝐞 𝐂𝐚𝐫𝐥𝐨 (MCMC) methods 🚀.

𝐌𝐚𝐢𝐧 𝐮𝐩𝐝𝐚𝐭𝐞𝐬:
✨ Renaming the project from PyMC3 to PyMC
✨ Replacing Theano with Aesara on the backend
✨ New JAX backend for faster sampling


𝐑𝐞𝐬𝐨𝐮𝐫𝐜𝐞𝐬 📚
Source code: https://lnkd.in/gue4HR25
Documentation: https://lnkd.in/gKj6sYZr
Release notes: https://lnkd.in/gVnf4Ktx

#bayesianstatistics #python #pymc #datascience #machinelearning