⦁ Anaconda
⦁ IPython
⦁ Python3 (+ modules)
⦁ Numpy
⦁ Pandas
⦁ Scikit-learn
⦁ Matplotlib
⦁ Seaborn
All these technologies are freely available and implementable. Therefore, the project is technically feasible.
Processor: Intel® Core™ i5-7200U @ 2.5 GHz
RAM : 4 GB DDR3
Hard Disk Space: 5 GB
Internet connection: To download the required software.
IDE: i) Anaconda 4.4.0
ii) Jupyter Notebook
Scripting Languages: IPython
Operating System: I) Windows 7 or newer
ii) MacOS 10.10+
iii) Linux (Ubuntu, RedHat, CentOS 6+)
- Clone the repository in your local Machine: git clone https://github.com/himanshu010/human-activity-recognition.git
- Change the directory into the cloned repository : cd human-activity-recognition
- Install the required python libraries : pip install -r requirements.txt
- Run the project in local Machine : Navigate & open Logistic regression.ipynb file in jupyter notebook
One area that has resonated greatly with activity recognition lately is sports, especially fitness and running. There are countless examples, in which the athlete or sportsperson is required to track his/her activity for selecting the next step in Training. But issues persist in the task to record his/her activity
People’s pleasure in tendency and to play never disappears, and there have been many improvements in the gaming area. But the problem is that humans cannot interact with the computers physically to get most of the entertainment of computer games. Only by recognizing the activities a user is performing, the computer can understand the user and give a response based on human reactions. This way, the human–computer interaction is possible.
The proposed system aims to predict the activity of humans by using Logistic Regression. This will help in recording the activity of athlete as well as facilitate the interaction between Humans and Computers. A brief overview of the functions is:
⦁ Exploring and cleaning the given data i.e., visualizing, dealing with null values, outliers, etc.
⦁ Applying a Logistic Regression algorithm to predict the activity based on the training data.
⦁ Getting rapid result by only selecting some of the important features that affect the prediction more.
The system takes the input data of the following files in CSV (comma separated values) format :-
⦁ X_train and Y_train (Values measured by sensors and their resulting Activity) as training data.
⦁ X_test and Y_test (Values measured by sensors and their resulting Activity) as testing data.
⦁ Take input values for Rapid Test
The system will generate the following outputs:
⦁ The system will output the visualization of the relation between activity predicted and actual activity.
⦁ Rapid Test Result and a plot showing its relationship with the ideal values.
1. The source code must be made public whenever a distribution of the software is made.
2. Modifications of the software must be released under the same license.
3. Changes made to the source code must be documented.
4. If patented material was used in the creation of the software, it grants the right for users to use it. If the user sues anyone over the use of the patented material, they lose the right to use the software.
