This Repository contains portfolio of data science projects that I have completed for self learning, and professional purposes. Presented in the form of Jupyter Notebooks and python files. For a more visually pleasant experience for browsing the portfolio, check out Gaurav
Tools
- Python: NumPy, Pandas, Seaborn, Matplotlib
- Machine Learning: scikit-learn, TensorFlow, keras
-
- K-Means Clustering - Mall_Customers.csv: Using K-Means Clustering on mall customers to make cluster on the basis of there spendings.
- Classification - Social_Network_Ads.csv: Using pipeline b/w classification model to find best model that classify that weather customer buy SUV or not.
- Advanced Regression - House Prices: Advanced Regression Techniques for prediction of house prices.
-
- Support Vector Machine - Spam_Classifier: Support vector machine that is used for classification of of spam or not.
- Passive Aggressive Classifier - Fake_news_classifier: Passive aggressive classifier which give pretty good accuracy in nlp that is used for classification of fake news.
- Random Forest Classifier - Stock Market Analysis: Random forest classifier to analyze stock market based on news heading sentiments.
-
- CIFAR-10: Transfer learning for classifing 10 different type of object i.e. ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck'].
- Xception - Face Mask: I have used Xception for classifing weather person is wearing mask or not. I have got accuracy near about 98% in validation data. Able to predict live using opencv.
- Convolution Neural Network - Digit Recognizer: Convolutional Neural Network that learns to recognize sequences of digits using data generated by concatenating images from MNIST.
- VGG16 - Fruits360: Transfer learning technique. I have used VGG16 for making classifier that classifiy 130 different types of fruits.
-
- Titanic Machine Learning from Disaster: Got 483 rank out of 23097 with 0.11682 score in Kaggle competition.
- Digit Recognizer: Got 1283 rank on building CNN model in Digit Recognizer compition.
-
- Car_Price_Predictions: Simple car price prediction trained on random forest regressor. Go and check out https://car-price-pred.herokuapp.com/
- Emotions_Analysis: Emotions analysis based on comments using natural language processing.
-
- House Prices Advanced Regression Techniques: Based off my advance house price analysis.
-
- Thanks to [Krish Naik (https://github.com/krishnaik06): Based off my advance house price analysis.
If you enjoyed what you saw, want to have a chat with me about the portfolio, work opportunities, or collaboration, feel free to contact me on:
* LinkedIn
* Instagram
* Twitter
* Website