📌 What is this tool?

This app https://aisciencetalk.blog/demo-tools/ lets you train and test AI models (multi-layer perceptrons) with multiple inputs and outputs using a friendly interface built in Streamlit. It’s ideal for:

• Regression-based tasks
• Educational demos
• Prototyping AI workflows on tabular data

💡 Why is this tool relevant?

Most AI tools are either too complex (requiring code) or too limited (black-box GUIs). This tool strikes a balance:

• You choose inputs/outputs
• 📈 You visualize model accuracy
• 🔍 You avoid extrapolation errors
• You make single-point predictions interactively

📁 Dataset Example: Measuring Speed

In this example, we use a dataset with:

• time – time taken to travel
• distance – distance covered
• speed – the output, calculated as speed = distance / time

time	distance	speed
[s]	[m]	[m/s]
1.0	3.0	3.0
2.0	6.0	3.0
4.0	8.0	2.0

The model will learn to approximate this relationship, even if the data includes some noise.

⚙️ How to Use the App

1. Upload your CSV

• The file should contain only numeric columns.
• You can use the example dataset: time, distance, speed.

2. Select input/output columns

• Inputs: e.g. time, distance
• Output: speed

3. Train the model

• The tool uses a shallow neural network with 2 hidden layers.
• Input values are scaled for better accuracy.
• Output includes:
  • R² scores (training & test set)
  • Scatter plot of predictions vs actual values

4. Predict new values

• Once trained, you can enter new inputs (e.g., time = 5.0, distance = 5.0)
• The tool will refuse predictions outside the training range to avoid unreliable extrapolation.

📈 Understanding the Results

• R² Score: Measures how well predictions match real outputs.
  • Closer to 1.0 = better fit.
  • Values < 0 = worse than just guessing the average.
• Prediction vs Real Scatterplot: Good models form a tight cluster around the diagonal line.

✅ Example Results (from our dataset)

Metric	Value
R² (Train)	0.999
R² (Test)	0.999
Prediction	time=5, distance=5 → speed ≈ 24.3

Need Something Similar?

This is a hands-on example of how artificial intelligence can be applied in real-world industrial engineering scenarios.

If you have sensor data (even raw), and want to extract value from it — let’s talk.

Contact us → https://aisciencetalk.blog/contact-us/

Disclamer

This content has been generated by an artificial intelligence language model. While we strive for accuracy and quality, please note that the information provided may not be entirely error-free or up-to-date. We recommend independently verifying the content and consulting with professionals for specific advice or information. We do not assume any responsibility or liability for the use or interpretation of this content.

Copyright

Author: Simone Togni

Platform: aisciencetalk.blog