This is the repository for the course Python for DSAI at Asian Institute of Technology.

Some resource worth mentioning:

1. Prerequisites/0 - Reading Roadmap

• For those who wants to know what papers to read. I have listed ONLY the most important papers you need to read in the field of machine learning

2. Prerequisities/0 - Installation

• For newbies who have trouble installing Python and other tools

3. Prerequisities/0 - Course Notations

• Understanding notations is the first step towards conquering math, so take a look and familiarized with it

4. Prerequisities/0 - Github

• As a data scientist, github is basically the most basic tool that you must know, so if you don't, please take a look

5. Syllabus/0. Course Introduction.ipynb

• Contains how I run the course. This course is a 15 weeks course, each week having two labs of 3 hours each. Each lab always end with the assessment and solution.

I would also like to give credits to several githubs that I have revised to create this:

• https://github.com/SethHWeidman/DLFS_code
• https://github.com/jakevdp/PythonDataScienceHandbook
• https://github.com/bentrevett/pytorch-sentiment-analysis

I would also like to thank students who have contributed:

• Akraradet Sinsamersuk
• Pranisaa Charnparttarvanit
• Chanapa Pananookooln

The course is structured into 3 big components, mostly focusing on preprocessing and modeling perspectives:

Focus on getting started.

• Python
• Numpy
• Pandas
• Matplotlib
• Sklearn

Focus on understanding the math + coding via coding from scratch

• Linear regression
• Polynomial regression
• Regularization

• Logistic regression
• Naive Gaussian
• Support Vector Machines
• Decision Trees
• K-Nearest Neighbors
• Bagging
• Random Forests
• Boosting - AdaBoost, Gradient Boosting

• K-means
• Gaussian Mixture Models

• Principal Component Analysis

• Linear Regression
• Logistic Regression
• Convolutional Neural Network

• Recurrent Neural Network
• Word Embeddings
• Padded Sequence
• Convolutional Neural Network for NLP
• Bidirectonal Stacked LSTM