The approach taken for this task is inspired by earlier work done in the year for Computer Vision (COMP3204) to approach this machine learning task. In the earlier work I came across [AutoGluon](https://auto.gluon.ai/stable/index.html#), an automatic machine learning library for Python. It simplifies model training using many different methods available and, where possible, it attempts to improve the model by itself.
Using this approach, for this part, it deduced that the best performing model, when looking at accuracy, was
| Model | Accuracy |
| ------------------- | -------- |
| LightGBM | 0.998333 |
| WeightedEnsemble_L2 | 0.998333 |
| LightGBMXT | 0.996667 |
| XGBoost | 0.995000 |
| RandomForestGini | 0.995000 |
| ExtraTreesEntr | 0.995000 |
| CatBoost | 0.991667 |
| LightGBMLarge | 0.990000 |
| RandomForestEntr | 0.990000 |
| NeuralNetTorch | 0.976667 |
| NeuralNetFastAI | 0.951667 |
| KNeighborsDist | 0.933333 |
| KNeighborsUnif | 0.905000 |
The chosen model in this case is `WeightedEnsemble_L2`
The calculated labels for this dataset was done sequentially, from `0` to `128` for the last column
This data was pulled directly from `TrainingDataBinary.csv`
As `WeightedEnsemble_L2` has an accuracy of 99.833%, I did not deem that it required any further tuning, and was therefore used for prediction to output `TestingResultsBinary.csv`
---
# Part B
This part used the same approach as in [[#Part A]]
| Model | Accuracy |
| ------------------- | -------- |
| LightGBMXT | 0.978333 |
| WeightedEnsemble_L2 | 0.978333 |
| LightGBM | 0.976667 |
| XGBoost | 0.973333 |
| LightGBMLarge | 0.970000 |
| ExtraTreesGini | 0.970000 |
| RandomForestGini | 0.968333 |
| ExtraTreesEntr | 0.968333 |
| RandomForestEntr | 0.966667 |
| CatBoost | 0.965000 |
| NeuralNetTorch | 0.881667 |
| NeuralNetFastAI | 0.876667 |
| KNeighborsDist | 0.856667 |
| KNeighborsUnif | 0.826667 |
For this part `WeightedEnsemble_L2` was also chosen as the goto model for prediction with 97.833% accuracy
### Python ###
# Byte-compiled / optimized / DLL files
__pycache__/
*.py[cod]
*$py.class
# C extensions
*.so
# Distribution / packaging
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
share/python-wheels/
*.egg-info/
.installed.cfg
*.egg
MANIFEST
# PyInstaller
# Usually these files are written by a python script from a template
# before PyInstaller builds the exe, so as to inject date/other infos into it.
*.manifest
*.spec
# Installer logs
pip-log.txt
pip-delete-this-directory.txt
# Unit test / coverage reports
htmlcov/
.tox/
.nox/
.coverage
.coverage.*
.cache
nosetests.xml
coverage.xml
*.cover
*.py,cover
.hypothesis/
.pytest_cache/
cover/
# Translations
*.mo
*.pot
# Django stuff:
*.log
local_settings.py
db.sqlite3
db.sqlite3-journal
# Flask stuff:
instance/
.webassets-cache
# Scrapy stuff:
.scrapy
# Sphinx documentation
docs/_build/
# PyBuilder
.pybuilder/
target/
# Jupyter Notebook
.ipynb_checkpoints
# IPython
profile_default/
ipython_config.py
# pyenv
# For a library or package, you might want to ignore these files since the code is
# intended to run in multiple environments; otherwise, check them in:
# .python-version
# pipenv
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
# However, in case of collaboration, if having platform-specific dependencies or dependencies
# having no cross-platform support, pipenv may install dependencies that don't work, or not
# install all needed dependencies.
#Pipfile.lock
# poetry
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
# This is especially recommended for binary packages to ensure reproducibility, and is more
