Table of Contents
Github Repo - https://github.com/Maimoons/7641/blob/master/README.md
Project 4: Reinforcement Learning The folder: Project 4 contains code for the forth assignment in the course. It implements 3 different RL algorithms on 2 different MDP:
- Policy Iteration- policy_iteration.py
- Value Iteration - value_iteration.py
- Q Learning - q_learn.py
- Experiments - experiments.py
Project 3: Unsupervised Learning
The folder: Project 3 contains code for the third assignment in the course. It implements 2 different clustering problems and 4 dimensionality reduction problems:
- K Means- kmm.py
- Expectation Maximization - em.py
- PCA - pca.py
- IPA - ipa.py
- Gaussian Random Mixture - grp.py
- Random Forest - rf.py
Project 2: Randomized Optimization
The folder: Project 2 contains code for the second assignment in the course. It implements 4 different problems:
- Neural Network- neural_network.py
- Knapsack - knapsack.py
- Traveling Salesman - tsp.py
- Continuous Peaks - continuos_peaks.py
Project 1: Supervised Learning
The folder: Project 1 contains code for the first assignment in the course. It implements 5 classifiers:
- Decision Trees- decision_tree.py
- Boosted Decision Trees - boost.py
- Neural Network - neural_network.py
- SVM - svm.py
- KNN - knn.py
Project 4 (the following on top of the dependencies from project 1)
- hiive.mdptoolbox
- mdptoolbox
- openaigym
Project 3 (the following on top of the dependencies from project 1)
- sklears.clusters
- sklearn.metrics
- sklearn.manifolds
- sklearn.mixture
- sklearn.decomposition
Project 2 (the following on top of the dependencies from project 1)
- MLrose --hiive
Project 1
- Numpy
- Pandas
- SkLearn
- Matplotlib
Follow the commands to setup and run the experiments:
- Clone the repo
git clone https://github.com/Maimoons/7641
The list of requirements for this repo is in requirements.txt as well as defined above so pip install each e.g
- pip
pip install numpy
There are two problems used in this project.
- Grid problem
- Non grid problem
The configuration for both is defined in the experiments file -
Running the experiments
python experiments.py 0
The output graphs are produced in the images folder.
There are two datasets used in this project.
- Dataset 0 - Titanic dataset from Kaggle
- Dataset 1 - Breast cancer dataset from sklearn
The dataset index needs to be passed in to run the experiments -
Running the experiments
python run_experiments.py 0
The output graphs are produced in the images folder.
The dataset used for the Neural Network problem -
- Dataset 0 - Titanic dataset from Kaggle
Running the classifier
python neural_network.py
Running any of the second half of optimization problems
python tsp.py
python knapsack.py
python continuos_peaks.py
The output graphs are produced in the images folder.
To plot training and testing time bar graphs for Neural Network -
python base.py
There are two datasets used in this project.
- Dataset 0 - Titanic dataset from Kaggle
- Dataset 1 - Breast cancer dataset from sklearn
The dataset index needs to be passed in to whatever classifier is being run. For example -
Running the classifier
python decision_tree.py 0
The above calls both the training and testing and saves the trained model. The output graphs are produced in the images folder.
To plot training and testing time bar graphs - Running the classifier
python base.py
Your Name - [email protected]