Letter Image Recognition Data with Machine Learning Algorithms
In this project, you need to classify 26 capital letters from 16 input values. The data set contains 20,000 samples, which are the features extracted from letter images.
Training data set: first 16,000 samples
Testing data set: the remaining 4,000
Download Dataset at http://archive.ics.uci.edu/ml/datasets/Letter+Recognition
- Choose at least three Machine Learning Algorithms for evaluation
- Use an existing standard Machine Learning algorithm toolkit, such as WEKA for your experiments.
- Evaluate the regression/classification performance / model size trade-off. Investigate the tuning potential by adjusting the parameters of your selected algorithms.
- Provide figures that explain your findings