Report of AI & Machine Learning Model For Step Data Analysis

The assignment focuses on developing a machine learning model to analyze step data from a wearable device for pet activity monitoring. Key tasks include data collection and preprocessing, model development, integration with firmware and app, and deployment strategy, with an emphasis on real-time analysis and optimization. Deliverables include a detailed report, Python code, a system architecture diagram, and a presentation summarizing findings.

Uploaded by

Bbb

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

18 views2 pages

Report of AI & Machine Learning Model For Step Data Analysis

Uploaded by

Bbb

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 2

Assignment: AI & Machine Learning Model for Step Data Analysis

Objective:

Develop a machine learning model that trains on step data received from a wearable device via
the Marshee app. The goal is to analyze step patterns and extract meaningful insights for pet
activity monitoring. Additionally, explore how this model can be integrated into the firmware and
app for real-time analysis and decision-making.

Assignment Tasks:

1. Data Collection & Preprocessing

● Explain how the step data will be collected from the device.
● Identify possible noise in the data and techniques for cleaning it.
● Discuss data normalization and feature engineering to enhance model performance.
● List any additional features that could be useful, such as activity duration, step
frequency, and rest periods.

2. Model Development

● Choose an appropriate machine learning model (e.g., Decision Tree, Random Forest,
LSTM, or CNN) and justify your selection.
● Implement feature selection techniques to improve model accuracy.
● Describe the training and validation process, including hyperparameter tuning.
● Discuss performance evaluation metrics (e.g., accuracy, F1-score, recall, precision) and
justify their use.

3. Integration with Firmware & App

● Outline how the trained model will be deployed on the Marshee Smart Tracker.
● Describe how real-time inference can be achieved on the device or via cloud processing.
● Explain how the app will display analyzed step data and provide insights to users.
● Discuss methods for optimizing model inference on an embedded system.

4. Deployment Strategy

● Define the pipeline for updating the model on the firmware through OTA (Over-the-Air)
updates.
● Suggest an architecture for cloud-based inference if real-time processing is required.
● Explain how periodic retraining will be handled based on new data.
5. Additional Considerations

● Highlight potential challenges in real-time processing and solutions.

● Discuss security and privacy considerations for step data.
● Propose potential AI-driven features such as step anomaly detection and fitness
recommendations for pets.

Deliverables:

1. A detailed report covering all the above tasks.

2. Python code or Jupyter notebook for model training and evaluation.
3. A system architecture diagram showing integration with firmware and app.
4. A brief presentation summarizing the key findings and implementation steps.

Evaluation Criteria:

● Data Processing & Feature Engineering (20%)

● Model Selection & Performance (30%)
● Integration Strategy (20%)
● Deployment & Optimization (20%)
● Clarity of Report & Presentation (10%)

Submission Deadline:

7 days after receiving

Note: Use relevant libraries such as TensorFlow, PyTorch, or Scikit-learn for model
development. The firmware deployment should be aligned with embedded AI best practices.