florianthom/dive-into-ml

Collection of small ml projects created at university. Contains jupyter-notebooks regarding e.g. required math, regression, ANNs, RNNs.

Dive-into-ML

This repository presents selected projects i did at university. The projects cover topics like exercises for required basic math, linear regression, logistic regression (classification), neural networks and recurrent neural networks and more.

Learned

• python
• anaconda
• jupyter + repl
• numpy
• tensorflow
• mxnet
• onnx
• Extended math knowledge
• Loss- and cost-functions
• Linear Regression
• Logistic Regression
• Bias-Variance-Tradeoff
• Decision Trees
• Gradient Descent
• Optimization algorithms (mostly gradient descent based)
• Feedforward neural networks
• Dropout
• Activation Functions
• Backpropagation
• Convolutions
• Convolutional neural networks (+ common CNN architectures)
• Recurrent neural networks
• LSTM / GRU
• Bidirectional neural networks
• Unitary neural networks
• Transformer networks