Machine Learning Notebook
This repository is tracking basic review from data analysis to advantage machine learning skills.
How to install ?
This repository depends on conda, so you must install conda at first, you can choose anaconda or miniconda. Then you should run bellow commands to setup a conda environment:
~ git clone git@github.com:classtag/machine-learning-notebook.git
~ cd machine-learning-notebook
~ conda env create -f environment.yml
~ conda activate machine-learning-notebook
~ ./run.sh
Basic package learning
- Numpy for linear algebra
- Pandas for data analysis
- Matpltlib for data visualization
- Seaborn for easier data visualization
Machine learning core algorithms
- Linear regression algorithm principle derivation
- Gradient descent strategy
- Logistic regression
- Decision tree algorithm
- Ensemble algorithm and random forest
- Bayesian
- Support vector machine
- Cluster KMeans
- Cluster DBSCAN
- Dimension reduction algorithm -PCA principal component analysis
- Neural network
- XGBoost
- Word2Vec
Case study and projects
- Python implements logistic regression and gradient descent strategies
- Abnormal transaction data detection
- Build decision tree model with scikit-learn
- Predict survival on the Titanic
- News classification task
- Tuning SVM
- Practice cluster alorithms
- Gensim library is used to construct the vector model of Chinese wiki baidu data words
- Scikit-learn modeling and evaluation
- Analyzes kobe's career
- Time series analysis
- Maximize profits from loan applications
- User loss warning
- EDA Football match dataset
- EDA Fao dataset
- HTTP log clustering analysis