Quant Pracar User Manual And Demo
Introducing Quant Pracar, the world's first educational application powered by Quantum Artificial Intelligence developed by Paanduv R&D. Designed to support both academic research and a wide range of industries, from healthcare to banking, Quant Pracar is a versatile tool for advancing quantum knowledge and applications. Our mission is to democratize quantum computing, ignite global enthusiasm, and establish India as a forefront leader in quantum computing and software development.
Key Features
Variety of Quantum-AI Algorithms: Choose from a diverse range of algorithms including QSVC, QNN, and more. Tailor your approach to suit your unique data and problem.
Intuitive Graphical User Interface: Dedicated User Interface specifically designed to Educate people during implementation with the help of flowcharts.
Training and Model Creation: Train your datasets with ease and create powerful models using our intuitive interface. Take control of your AI journey and harness the potential of quantum algorithms.
Hyperparameter Tuning: Fine-tune your models with precision using our hyperparameter tuning capabilities. Optimize performance and unlock new insights from your data.
Real-Time Quantum Circuit Visualization: Witness the magic of quantum computing in action with our real-time quantum circuit visualization. Gain a deeper understanding and insight into the inner workings of quantum algorithms.
Dedicated Prediction Window: Make predictions effortlessly with a dedicated prediction window. Streamline your workflow and make informed decisions with just one click.
Launching Quant Pracar
Once Quant Pracar is initiated, the terminal will open in the background along with the GUI.
Home Page Overview
Upper Right Corner: The logo for Quant Pracar.
Center: A flowchart guiding you through the current process.
Navigation Bar: This allows you to choose the module you want to access first.
Modules Provided by Quant Pracar’s GUI
Train Model: Allows users to train a model using their data for various purposes.
Prediction: Allows users to make predictions using their pre-trained model.
Here as we are learning from scratch we will first go to the Train model module and once training is completed we will proceed with prediction.
Launching new project
Step 1: Create a New Project
Enter the project name in the entry box.
Click the “Create New” button and browse to the directory where you want to save your project information.
Step 2: Upload Dataset
After creating the directory, click the “Upload Dataset” button.
Ensure your dataset is in CSV format, with the last column being the dependent variable.
Step 3: Select Algorithm
Choose an algorithm from the available options.
Click the “Save and Proceed” button to move to the next step.
Model Training
Step 1: Select the Feature Map and Algorithm
Begin by choosing a Pauli feature map and the algorithm you wish to use.
Step 2: Set Hyperparameters
Click the button in the upper left corner to select the algorithm and proceed to the “Hyperparameters” step.
Choose the hyperparameters for training your model.
Save your settings and click the “Summary Report” button to continue.
Step 3: Review and Train the Model
The summary report will display all selected parameters for review.
Confirm your selections and start the model training.
Upon completion, a “Training is completed” message will appear below the button.
Step 4: Monitor Training and Evaluation
Monitor training progress in the terminal.
Once training is complete, the model will be evaluated, and results will be available in the evaluation window.
Step 5: View Evaluation and Visualization
Alongside the evaluation report, access the visualization window to view the loss function and Quantum circuit.
Step 6: Access Project Files
Navigate to the project folder to find various files, including:
model. pkl: Contains training information.
Text files: Include the summary report and parameter reports.
Preprocessed data: Useful for labeling data for prediction.
Plots: Visual representations of the loss function and Quantum circuit.
Making Predictions
Step 1: Access Prediction Window
Ensure you have a trained model and its summary report.
Select the algorithm on which the model was trained.
Choose the required parameters.
Step 2: Upload Model and Input Data
Upload the path to the trained model and the input data for prediction.
Click the “Predict” button.
Step 3: View Prediction Results
Once the prediction is complete, the results will be displayed below the button.
Check the project folder for the prediction.csv file to view the predicted values.