Ensure your data is clean before uploading for best results.
Tip: If you encounter a "Error reading file", double-check the file format and ensure it's not corrupted.
2. Review Dataset Previews
The "🔍 Original Dataset Preview" shows your data as uploaded.
The "🔄 Updated Dataset Preview" reflects changes after operations like missing value handling.
Data Preprocessing
Missing Values Report & Handling
The "📊 Missing Values Report" on the right provides a summary of missing data.
Automatic Handling: Choose a method (Mean, Median, Mode, Forward/Backward Fill) to apply to all relevant missing values. This is great for a quick fix.
Manual Handling: Select specific columns with missing values and choose a precise imputation method (Mean, Median, Mode, or custom value) for that column.
Important: Handling missing values is crucial for accurate model training. Numerical columns can be filled with Mean/Median, while categorical with Mode or a custom string.
Data Operations & Algorithms
1. Define Target Variable and Problem Type
Located in the top-left expander "⚙️ Data Operations & Algorithms".
Select the column your model will predict (the 'Y' variable).
Visio AI will try to auto-detect if it's a Classification (predicting categories) or Regression (predicting numbers) problem.
2. Split Data into Train and Test Sets
After defining the target, you can split your data.
Adjust the "Test Set Size" slider (e.g., 0.2 means 20% for testing).
Set a "Random State" for reproducible splits.
Note: Visio AI automatically handles basic encoding of categorical features and scaling of numerical features during this step.
3. Select Machine Learning Algorithm
Based on your problem type (Classification/Regression), a list of suitable algorithms will appear.
Choose the algorithm you want to train.
Tip: Different algorithms suit different data types and problem complexities. Experiment to find the best fit!
4. Train Model & View Results
Click the "🚀 Train Model" button.
The "🧠 Machine Learning Operations" section at the bottom of the page will display performance metrics (Accuracy, MSE, R2, etc.) and a Confusion Matrix for classification.
Visualization
Generate Pair Plot
Use the "📈 Generate Pair Plot of Numerical Columns" button below the Missing Values Report.
This creates a grid of scatter plots for all numerical column pairs, helping you visualize relationships.
Both static (Seaborn) and interactive (Plotly) plots are generated.
Tools in Sidebar
📝 Note -- Lite: For quick notes and text editing.
😶🌫️ WordCloud: Generate word clouds from text data (requires text input).
🤖 Viz AI (img): (Future feature) AI-powered image visualization.
🧮 Calculator: A simple calculator.
⚙️ Viz Editor: (Future feature) Advanced visualization editing.
📄 Viz Report: (Future feature) Generate a comprehensive data analysis report.