Yiqiao Yin

WYNAssociates. This is the official AI-driven solution offered by W.Y.N. Associates. WYN aims to build the best data science lab with the assistance of smart brains and AI-driven technologies for our local and global clients. With the advancement of WYN’s Explainable Artificial Intelligence (XAI) technology, we drive a set of processes and methods that allows human users to comprehend and trust the results and output created by machine learning algorithms.

Introduction-to-Machine-Learning-Big-Data-and-Application. This github repo introduces a portfolio of probability theory, statistical inference, and applied machine learning topics.

Statistical-Machine-Learning. This is the Github repo for the field of Statistical Machine Learning. I set this up as my personal blog for future generations and for anybody who is interested. Please feel free to contact me on LinkedIn if you have questions.

huggify-data. This is the repo for huggify-data package.

deepspeed-course. Accelerate deep learning with DeepSpeed: distributed training, multi-GPU examples, and practical guides for neural nets and HuggingFace models.

sat-locat-reference-team-lucc. The teamlucc package is designed to facilitate analysis of land use and cover change (LUCC) around the monitoring sites of the Tropical Ecology Assessment and Monitoring (TEAM) Network. The TEAM Network is a global network of sites in tropical forests wth standardized real-time data collection designed to measure tropical forest responses to climate variability and change, land cover and land use change, and other threats.

fine-tune-openai-gpt-oss. Fine tuning script using deepspeed.

recurrentshop-rnn-reference. Ability to easily iterate over different neural network architectures is key to doing machine learning research. While deep learning libraries like Keras makes it very easy to prototype new layers and models, writing custom recurrent neural networks is harder than it needs to be in almost all popular deep learning libraries available today. One key missing feature in these libraries is reusable RNN cells. Most libraries provide layers (such as LSTM, GRU etc), which can only be used as is, and not be easily embedded in a bigger RNN. Writing the RNN logic itself can be tiresome at times. For example in Keras, information about the states (shape and initial value) are provided by writing two seperate functions, get_initial_states and reset_states (for stateful version). There are many architectures whose implementation is not trivial using modern deep learning libraries, such as:

aws-deploy-deepseek-from-scratch. This repo is a tutorial to walk through steps to deploy deepseek-r1 on AWS from scratch to API.

pace-u-cs667. This is course repo for Practical Data Science. https://yiqiao-yin.github.io/pace-u-cs667/. If you are prepping for Capstone, please use the following link.

Applied-Statistics. This repo preserves the knowledge of Applied Statistics by Yiqiao Yin, a PhD student at Columbia University. I would like to acknowledge this repo is open source and non-profit only. In addition, I want to thank professors who instructed this course to provide the knowledge and resources for us.

Fundamentals-in-Machine-Learning. This is the Github repo for the field of Statistical Machine Learning. I set this up as my personal blog for future generations and for anybody who is interested. Please feel free to contact me on LinkedIn if you have questions.

sat-locate-reference-deepOSM. Classify roads and features in satellite imagery, by training neural networks with OpenStreetMap (OSM) data.

googleAuthR-reference. Auto-build libraries for Google APIs with OAuth2 for both local and Shiny app use. This guide is also available at the googleAuthR website The googleAuthRverse Slack team has been setup for support for using googleAuthR and the libraries it helps create. Sign up via this Google form to get access.

matlab-cnn-reference. Matlab codes for 2D Convolutional Neural Network

Communicating-Statistics. This is the github repo for related projects from Communicating Statistics course by Professor Andrew Gelman

YinsPy. This is repo for a list of small python projects that I found interesting.

wyn-chatbot-io. This chatbot stores sample use cases developed using Streamlit, ChatGPT, and Palm2. It allows users to interact with LLM via multiple different use cases.

grok-cli. This repo enables Grok-CLI from command line.

aws-textract-tutorial-2025. Python

Fundamentals-of-Reinforcement-Learning. Jupyter Notebook

chainlit-mcp-sql-demo. This repo demo a chainlit UI with mcp invoking SQL agent.

YinsPredictor3_0. This package is designed to assist money managers to build better understanding of action-based trading behaviors.

YinsKerasNN. This package uses Keras as framework which provides a high-level neural networks API developed with a focus on enabling fast experimentation.

WYN-Langchain-DocQuery. Python

WYN-ESL. This is the AI tester backed by ChatGPT.

YinPortfolioManagement. This package uses Long Short-Term Memory (LSTM) to forecast a stock price that user enters.

matlab-deep-learn-reference. Deep Learning is a new subfield of machine learning that focuses on learning deep hierarchical models of data. It is inspired by the human brain's apparent deep (layered, hierarchical) architecture. A good overview of the theory of Deep Learning theory is Learning Deep Architectures for AI

Image-Process-Yelp. Deep Learning using MXnet and Inception-BN in R

MNIST_Data_Using_CNN_and_Tuning. This is the R code for tuning CNN using MINST Fashion data set. The CNN code contributed includes the following contents:

retina-analysis-reference. Biologists who study vision are fascinated by the retina, the rear surface of the eye which collects light and transmits an 'image' to the brain. Through evolution, eyes have been exceptionally diverse across the animal kingdom, yielding a wide variety of shapes, sizes, colors, and chemical processes. For this reason, the eye is often used to understand how complex traits evolve with ecology, the interactions of the animal with its environment. A retinal ganglion cell map helps biologists visualize receptive cells across the rear surface of the vertebrate eye. This topographic map shows density in a way which highlights areas of higher visual performance, and this can be used to understand what regions of an animal's visual field are most clear-in-view. For a human, we only have one point on the eye where things are clear- that's why we can really only see a few words at a time while reading, for example. Some fish have multiple centers of high visual performance, more formally referred to as a peak of retinal acuity.

neuraltalk-rnn-reference. Warning: Deprecated. Hi there, this code is now quite old and inefficient, and now deprecated. I am leaving it on Github for educational purposes, but if you would like to run or train image captioning I warmly recommend my new code release NeuralTalk2. NeuralTalk2 is written in Torch and is SIGNIFICANTLY (I mean, ~100x+) faster because it is batched and runs on the GPU. It also supports CNN finetuning, which helps a lot with performance.

pace-u-cs676. This is the course repo for Algorithms in Data Science.