Option challenge : Options Explorer

Imagine that you started working in a company that collects and analyzes high-frequency grinding machine sensor data to detect machine anomalies. You are tasked to find an optimal model for your dataset under certain limitations and preferences.

The intent can be described as anomaly detection with the general method of analysis multiclass classification. Preferably, your model should fulfill the following soft constraints: (1) accuracy of at least 0.9, (2) precision of at least 0.9, (3) recall of at least 0.8, (4) F1 score of at least 0.85, and (5) the processing unit (PU) to be GPU. 

Prerequisites

No other challenge has to be completed prior to this one.

Background

The Options Explorer provides an optimal strategy for data analysis under user limitations and preferences. The user wants to find an optimal model for his specific problem. He/she provides the background story of the dataset, e.g. domain of knowledge, intent, and general method of analysis, as well as the hard and soft constraints. The probabilistic model based on the Markov decision process recommends the three best ranked models.

Exercise

Complete the following steps to solve the challenge.
1. Download

the dataset UC5_expanded.csv from OptionsExplorer materials

2. Open

the Options Explorer GUI http://194.249.3.27:3000/

3. Register
4. Click on Create an experiment and then CSV Upload.

Select the dataset UC5_expanded.csv and upload it. 

5. Click on Explorer and provide the context.

In Required Parameters, choose the domain of knowledge as the manufacturing industry. In Optional parameters, specify the intent as anomaly detection and the general method of analysis as multiclass classification. They will serve as hard constraints. 

6. Click on Add Constraints and specify the soft constraints.

Choose the variables from the drop-down menu and specify thresholds or categories for each of them. 

  • (1) accuracy ≥ 0.9, 
  • (2) precision ≥ 0.9
  • (3) recall  0.8, 
  • (4) f1_score ≥ 0.85, and 
  • (5) pu = GPU. 
7. Click on Run the MDP Process.
8. Check out the results in the ranked list of the models.

Success criteria

Make a screenshot of the ranked list of the best three models and upload it to
the Options Explorer Solution Screens.hots folder
The ExtremeXP project is co-funded by the European Union Horizon Program HORIZON CL4-2022-DATA-01-01, under Grant Agreement No. 101093164

© ExtremeXP 2026. All Rights Reserved – Privacy Policy