arslan-rashidov / clients-interviews-analyze

Clients Interviews Analyze

API

1. cd to the folder
2. python3 -m pip install -r requirements.txt
3. python3 -m ApiController
4. http://127.0.0.1:8000/docs

Used Libraries

1. Uvicorn
2. FastApi

Model Building

Used libraries

1. Pandas
2. Numpy
3. Sklearn

Data Preprocessing (data_preprocess.py)

Our current dataset consists of:

1. Main Technology of a candidate
2. Seniority level of a candidate
3. English level of a candidate
4. Location ID of a candidate

Features were prepared with OneHotEncoder.

Analyze (interviews_analyze.py)

Train - 80% Test - 20%

Best parameters were chosen with GridSearchCV.

Algorithms

Random Forest accuracy - 85.2%