Learn about the following terminologies:
- Accuracy, Precision and F-Score
- Curse of Dimensionality
- Bias Variance Tradoff
- Loss Functions
- Visual Introduction to Machine Learning Part 1
- Visual Introduction to Machine Learning Part 2
- Introduction to Curse of Dimensionality
- UC Davis Introduction to Genetic Algorthms
- Evolutionary computation course (AEC 02 and 03 only)
- https://lethain.com/genetic-algorithms-cool-name-damn-simple/
Implement a GA to extremize a polynomial within range.
Youtube videos - good, intuitive, in-depth https://youtu.be/aircAruvnKk
http://neuralnetworksanddeeplearning.com/
- https://youtu.be/aircAruvnKk
- http://neuralnetworksanddeeplearning.com/
- http://playground.tensorflow.org
There are 11 TODOS in this python file Fill each one of those appropriately and you will have a working neural network Instructions and resources have been provided wherever possible. The implementation may not be perfect, so feel free to point out any mistakes / ask any doubts
After completing the task, some of the things you could try are (optional):
- Implement different cost functions (binary cross-entropy)
- Implement different activation functions (tanh, ReLU, softmax)
- Incorporate these changes in the neural netwok code so that you can select the loss / activation function
- Play with the hyper-paramters!
Combining Genetic algorithms and neural networks Read up on these resources as. Later we will ask you to implement something similar.