Project delves into implementing the K-Nearest-Neighbor Algorithm by fitting and predicting target attributes via making testing and training data sets for k-fold cross validations and 1-NN Algorithims using Python libraries and custom Python classes!
While we will be working with the k-Nearest Neighbor algorithm in various ways, we will also implement methods for data pre-processing, debugging, model evaluation, normalization, standardization, and visualization to create, evaluate, and optimize a data mining algorithm as applied to a real-world dataset.
Project Utilizes:
- Pandas
- Series/Dataframes
- Loaded Operators
- Higher Order functions
- NumPy
- linalg module
- SciPy
- distance module
- Sklearn
- BaseEstimator module
- ClassifierMixin module
- Timeit
- default_timer() method
- Visualization
- seaborn module
- matplotlib.pyplot module
- Python kNN Algorithim
- sklearn.neighbors module
Project Features:
- Project delves into several different ways to implement the 1-NN Algorithim
- Project delves into several different ways to implement the K-Nearest-Neighbor Algorithm
- Different algorithms are timed to help determine which method of implementing the K-Nearest-Neighbor Algorithm is the most efficient.
- Algorithm plotted visualizations to see the impact on the algorithm's duration and efficiency.