Prototype Models Example

Prototype Models Example

Problem Statement

Use the Neural Network model to predict the values for gear based on the other variables.

How to perform analysis

Step 1: Open Sigma Magic
  1. Click on the Sigma Magic button on the Excel toolbar.
  2. Click on the New button to create a new project.
Step 2: Add the analysis template
  1. Click on the Tool Wizard to add the analysis template.
  2. Click on Analytics and then Prototype Models.


Step 3: Specify analysis options
A new worksheet will be added to your workbook. Analysis Setup will be automatically opened, in the setup tab specify the survey results.


Click on Data to specify the data required for this analysis.   


Click on the Train the software will let you pick the options for training the given model. Training is a step where we split the data into groups a train data set and a test data set.


Click on the Tuning to identifying the best set of hyperparameters that gives the best fit for the given model.


Click the Verify tab to ensure all the inputs are okay and shown in a green checkmark. 


Step 4: Generate analysis result
Click OK and then click Compute Outputs to get the final results.



Interpretation of Results

  • The chosen prototype model is k-NN, a classification algorithm that predicts outcomes based on the closest neighbors in the dataset.
  • The accuracy of the model is 73.73%, meaning that roughly 74 out of 100 predictions are correct.
  • Objective: The model is used for classification.
  • Data Features Used: Variables like x1, x2, x3, x4, x5, etc.
  • Hyperparameters Tuned:
    • Tuning Length: 3 (suggesting the model was optimized by testing 3 different configurations).
    • Tuning Method: Random Search to find optimal parameters.
    • Distance Metric: Likely Euclidean or Minkowski distance, as commonly used in k-NN models.
  • Training Data Size: 80 rows, divided into multiple classes.
  • Resampling Method Used: Likely cross-validation to avoid overfitting.
  • Model Predictions Table:
    • The model predicts values based on the nearest neighbors.
    • Predictions are influenced by input variables (x1 to x5).
  • Accuracy Scores for Different Runs:
    • Scores range between 0.5735 to 0.6553, showing variance in performance.
    • The final model's best accuracy is 73.73% after resampling.
  • Validation Method: Likely out-of-bag error validation, ensuring robustness.
  • A Variable Importance Plot is present, showing which input variables had the most influence.
  • Key influential features seem to be x2, x3, x4, which significantly impact the prediction outcome.
  • The model performs decently but has scope for improvement.
  • The next steps could be:
    • Tuning hyperparameters further, such as adjusting k values or using different distance metrics.
    • Trying alternative models, such as Decision Trees or Random Forest, to compare accuracy.
    • Feature selection & data preprocessing to improve prediction reliability.

    • Related Articles

    • Prototype Models Overview

      Prototype models serve as foundational frameworks for developing and testing analytical solutions. These models help organizations gain insights, make data-driven decisions, and improve processes. Below is an overview of key prototype models in ...

    • Prototype models frequently asked questions

      What is a Prototype Model? A prototype model refers to a preliminary version of an analytical model developed to test hypotheses, validate assumptions, and refine processes before full-scale implementation. These models help in rapid experimentation ...

    • Regression Models Example

      Problem Statement Determine a regression model between input X (advertisements) and Y (sales). The data for this exercise is given in the Data tab. Fit a regression model between the input(s) and the output? Is this model statistically significant? ...

    • Bayesian Models Overview

      Bayesian models are a class of statistical models based on Bayesian probability, which provides a probabilistic framework for updating beliefs based on new evidence. These models use Bayes' theorem to update prior knowledge with observed data to ...

    • Boosted Models Overview

      Boosted models are a class of ensemble learning methods used in machine learning to improve predictive accuracy by combining multiple weak learners (typically decision trees) into a strong model. These models are built sequentially, where each new ...