Create the best possible neural net i.e. the one with maximum number of layers and fewest number of parameters that will solve the XOR problem. Give the weights and intercepts for each neuron and other parameters that you have used.
Create a neural net for a dataset chosen from the UCI ML repository. The repository is available at: http://archive.ics.uci.edu/ml/datasets.html
- You will first have to read in the dataset using Pandas into a dataframe.
- The second step will involve pre-processing the dataset - analyze each of the attributes and scale them.
- Then you will randomly split the data into train and test parts – you are free to decide the split size. Next will be the model creation step – you will need to tune as many parameters as possible.
- Finally, evaluate the performance of the model using the best set of parameters.
- For part 2, I have used the dataset from this url => https://archive.ics.uci.edu/ml/machine-learning-databases/iris/
- Iris dataset : https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data
- Iris dataset column names => https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.names
Before attempting this lab, code snippets from following link were referred https://elearning.utdallas.edu/bbcswebdav/pid-1826701-dt-content-rid-22254227_1/courses/2182-UTDAL-CS-6375-SEC003-24048/code.txt