spaceface02 / compendium_ml_projects Goto Github PK

1. Anaconda Distribution, with Jupyter Lab/Notebook along with libraries like Pandas, Matplotlib, Seaborn, Numpy, Tensorflow, OpenCV, Sci-Kit Learn, PIL, NLTK, Spacy, Plotly, Cufflinks etc.
2. Make virtual environments for each of the projects, and install the pacakges mentioned in te import statements.

• python -m venv venv
• source venv/bin/activate
• pip install -r requirements.txt
• python file.py or select the virtual env from jupyter notebook/lab.
• If using an IDE or code editor, be sure to change the python interpretor path to the python executable in the venv folder.

P.S If the source command doesn't work, then do the following in place of that:

• cd venv\Scrips\activate.bat

• Blue Pen Tracking - Tracks a pen with a blue cap across the videocam, bu drawing streaks of color.
• Book Reviews - Using a dataset from Kaggle, find the genre of book based on various features.
• Cartoonify Yourself - Cartoonify your own image using a pretrained model.
• Detecting Fake News - Detecting fake news using a Passive Aggressive Algorithms. (References present in the notebook).
• Drowsiness Detector - Detects if the user is drowsy or not in a car while driving.
• Ensemble, Forests, Trees - Exploring various tree models and their performance.
• Eye Dropper Color Picker Tool - Detects any color in any image, works just like a color dropper in Photoshop.
• Human Detection - Detects human in a video using HOG Descriptor. (References present in the notebook).
• Lane Detection - Detects the lanes in a video. Uses Hough Lines. (Reference video present in the notebook).
• License Detector - Detects the license plate in an image.
• Multiple Color Detection - Detects the colors in a live stream video.
• Netflix Movie Reviews - Exploratory Data Analysis on the Movie Reviews dataset on Kaggle. Also building a content and collaborative filtering recommender system.
• Parkinson Disease Detection - Detects if a person is suffering from Parkinson's disease using a datset from Kaggle
• Quora Question Pairs - Removing the duplicate questions from the Quora dataset from Kaggle. Done for the IECSE chatbot project.
• Time Series Analysis (RNN's, LSTM's) - Predicting and drawing the sine wave from previous data and predicting sales forecase for a clothing company (RSCCASN).
• Smile Selfie Taker - Takes a picture only when the user is smiling.
• Student Marks - Doing Explorative Data Analysis on the Student Marks dataset from Kaggle.
• Support Vector Machines - Exploring SVM's in more detail.
• Titanic - Exploring the Titanic dataset from Kaggle. Also using Plotly and Cufflinks to plot interactive graphs.
• Uber Project - Doing some EDA on the Uber dataset from Kaggle. Detecting location of the users calling cabs, frequencies of cabs and other interesting aspects.
• Wine Quality Prediction - Exploring the Wine Quality datasset from Kaggle and classifying wine in 5 classes, based on their quality, from various features, like density, citric acid content, chlorides etc.
• Word Clouds - Generates a word cloud from a text file. In this case, the text comes from the youtube comments of the music videos of Eminem.