/Web-Data-Mining-Project-2021-SCU-

In this work, To implement this project we used Python Programming language and Libraries: Scikit-Learn/SciPy, NumPy, Pandas, Matplotlibwe carried out and evaluated a classification algorithm to notice 4 fundamental human physical activities(walking, cycling, sitting, and lying) using five triaxial accelerometers worn simultaneously on unique parts of thebody (dominant hip, higher arm, ankle, thigh, and wrist). The accelerometer information was collected, cleaned, and preprocessed to extract elements from the 10 s window. These time and frequency area aspects had been used with Random Forest and k-Nearest Neighbour classifier to classify challenge activities. The algorithms had been evaluated based totallyon Leave-One-Subject-Out (LOSO) and ten-fold cross-validation method the usage of each accelerometer records as well as annotated undertaking labels from 33 contributors in a lab. Random Forest showed the first-rate overall performance recognizing the activities with common accuracy of 89 % for the LOSO approach for hip data. Combining statistics from both hip and ankle improved the common accuracy with the aid of 3.5 %, and by using 10% for mendacity activity, which had the lowest classification accuracy (80%) for hip data. We conclude that our algorithm that makes use of 10 aspects suggests right endeavor classification, and is computationally efficient to be carried out in real-time cell systems.

Web-Data-Mining-Project-2021-SCU

#This project is done for academic purpose only. #webdatamining, #Machinelearning, #Python #SCU

*Team Leader:Chandan Kumar Sah(2018511460047)

*Team Members:

1.Sudin Upadhaya(2018511460046) 2. Mir Parsin Fatema(2018511460031)

Data-Mining

Recognizing human activity using multiple wearable accelerometer sensors placed at different body positions.

This project was created as a part of Data Mining course at Sichuan University, China.

In this project, we implemented and evaluated classification algorithm to detect four crucial human physical activities (walking, cycling, sitting, and lying) using five triaxial accelerometers worn concurrently on different parts of the body (dominant hip, upper arm, ankle, thigh, and wrist). The accelerometer data were collected, cleaned, and preprocessed to extract features from 10 s window.

These time and frequency domain features were used with Random Forest and k-Nearest Neighbour classifier to classify subject activities. The algorithms were evaluated based on Leave-One-Subject-Out (LOSO) and ten-fold cross-validation strategy using both accelerometer data as well as annotated activity labels from 33 participants in a lab.

Random Forest showed the best performance recognizing the activities with overall accuracy of 89 % for LOSO strategy for hip data. Combining data from both hip and ankle improved the overall accuracy by 3.5 %, and by 10% for lying activity, which had the lowest classification accuracy (80%) for hip data.

  • Technologies Used to Implement this project:
  • Language: Python programming
  • Tools: IPython, PyCharm IDE
  • Libraries: Scikit-Learn/SciPy, NumPy, Pandas, Matplotlib
  • Algorithms : Random Forest, K-NN
  • Evaluation Methodology: LOSO(Leave One Subject Out), K-Fold validation, Confusion Matrix

Abstract:We have finished a amazing course with our respected teacher. He taught us very well about Web data mining. Now, this is our turn to show what did we learn from our teacher. This report included with our project,the codes of our system,the processes to implement our data sets,what we have learnt about ,Machine learning and Web Data Mining.In this work, To implement this project we used Python Programming language and Libraries: Scikit-Learn/SciPy, NumPy, Pandas, Matplotlibwe carried out and evaluated a classification algorithm to notice 4 fundamental human physical activities(walking, cycling, sitting, and lying) using five triaxial accelerometers worn simultaneously on unique parts of thebody (dominant hip, higher arm, ankle, thigh, and wrist). The accelerometer information was collected, cleaned, and preprocessed to extract elements from the 10 s window. These time and frequency area aspects had been used with Random Forest and k-Nearest Neighbour classifier to classify challenge activities. The algorithms had been evaluated based totallyon Leave-One-Subject-Out (LOSO) and ten-fold cross-validation method the usage of each accelerometer records as well as annotated undertaking labels from 33 contributors in a lab. Random Forest showed the first-rate overall performance recognizing the activities with common accuracy of 89 % for the LOSO approach for hip data. Combining statistics from both hip and ankle improved the common accuracy with the aid of 3.5 %, and by using 10% for mendacity activity, which had the lowest classification accuracy (80%) for hip data. We conclude that our algorithm that makes use of 10 aspects suggests right endeavor classification, and is computationally efficient to be carried out in real-time cell systems.

Keywords:Web Data Mining, Machine Learning,Python,Scikit-Learn/SciPy, NumPy, Pandas, Matplotlib, Activity recognition, Random Forest, k-NearestNeighbors, leave-one-subject-out,

#How to Run the project All the steps of web data mining are completely done using python programming language-2021 *Libraries Dependencies: pandas numpy scipy sklearn matplotlib pylab

  • Install dependencies, steps below: *Open cmd
    • Goto C:\Python27\Scripts (In windows) *Run "pip install pandas" *Run "pip install numpy" *Run "pip install scipy" *Run "pip install sklearn" )Run "pip install pylab"(Note: After you install numpy, scipy and Sklearn then only Pylab will be install) *(Note: I can't share my personal project report here, If annyone need help to write it kindly follow the above table of contains or Contract me through my Linkedin Profile ID :https://www.linkedin.com/in/chandan-kumar-sah-5803bb1b4/ )

Thanks! Have a good luck

Regards!

Chandan Kumar Sah