/Optimal-Classification-Model-of-BLE-RSSI-Dataset

Development optimal classifiers to predict the positions of the user based on the RSSI readings from iBeacon devices

Primary LanguageHTMLMIT LicenseMIT

OPTIMAL CLASSIFICATION MODEL OF BLE RSSI DATASET FOR INDOOR LOCALIZATION AND NAVIGATION

Project Overview

The aim of the project is to develop optimal classifiers to predict the position of the user based on the RSSI readings from thirteen iBeacon devices. The BLE RSSI labeled dataset was used to train different classifiers and evaluate their performance.

Motivation

The motivation for the project is to develop big data analysis tool for smart cities and help people guide indoors using machine learning approach.

Bild Status

Scrutinizer

CocoaPods

Tech/Framework used

The project was done using python GUI Anaconda.

Features

  • Logistic Regression
  • Decision Tree
  • Random Forest
  • K-fold CV
  • GridSearchCV
  • Pipeline
  • PCA
  • SelectKbest
  • Voting

Dataset

  • Labeled dataset -Number of variables: 15 -Number of data points: 1420
  • Unlabeled dataset -Number of variables: 15 -Number of data points: 5191

Code Example

#The following code predicts the best outcome of the input
Prediction = voting_clf.predict(Input)

References

The following book and python machine learning library manuals were followed to complete the projcet. The code was not directly copied from the following referances without proper inline citation.

License

MIT @ JBP261