ML-and-DL-with-TF

Machine learning and deep learning algorithms and techniques using tensorflow and sklearn

Regression_Analysis:

Full linear regression ml journey, with data exploration, visualization, preprocessing, algorithm implementation, model comparison, and feature selection.
Linear regression and Random Forest regressor performance evaluation, comparison and feature selection on california housing prces dataset.
Conclusion: Random Forest Regressor with only 4 features out of 8 performed better.

Batch_Normalization_CNN:

CNN Model performance evaluation with different activation functions and effect of batch normalization.
CNN model for fashion mnist data, and comparison of relu and sigmoid activation functions with added effect of batch normalization.
Conclusion: Batch Normalization improved performance for both activation functions and Relu performed better.

RNN:

Simple RNN for addition using strings as inputs.

Deepfake_with_keras:

Implementation of Deep Convolutional Generative Adversarial Networks (DcGan) for fenerating synthetic images of numbers(0-9) by traing the GAN on mnist data.