Welcome back (previous post<\/a>) to my machine learning journey! In the last post, I explained the setup environment and demonstrated the first steps of building an interactive machine learning model. Today, I’ll attempt to (with my limited knowledge) explain creating the web application using Flask<\/a>. It serves as an example of the many ways you could interact with a model in real-time.<\/p>\n\n\n\n Flask is a lightweight web framework for Python that’s perfect for building simple web applications quickly and efficiently. It provides the tools needed to set up routes, handle requests, and render templates.<\/p>\n\n\n\n To get started with Flask, you’ll need to install it:<\/p>\n\n\n\n Once installed, you can create a basic Flask app by defining routes that handle different actions.<\/p>\n\n\n\n The core functionality of our web app revolves around querying instances from the pool and allowing users to label them. Let’s set up the necessary routes and create HTML templates for user interaction.<\/p>\n\n\n\n Here’s the basic structure of our Flask app:<\/p>\n\n\n\n One of the critical features of our app is the ability to correct labels. This ensures that any mistakes made during the labeling process can be rectified, improving the model’s learning over time.<\/p>\n\n\n\n In the Creating the web application presented several challenges, from handling data correctly to ensuring that the user interface was intuitive and responsive. Here are some of the key challenges I faced and how I overcame them:<\/p>\n\n\n\n Reflecting on this part of the journey, I realized the importance of building a user-friendly interface and the need for careful data handling. These elements are essential for creating a robust interactive machine learning model.<\/p>\n\n\n\n In the next post, we’ll explore how visualizing data can help us understand our model better. Stay tuned!<\/p>\n
\n\n\n\nIntroduction to Flask<\/h3>\n\n\n\n
bashCopy codepip install Flask<\/code><\/code><\/pre>\n\n\n\nCreating Routes for Querying and Labeling Instances<\/h3>\n\n\n\n
pythonCopy code
import numpy as np\nimport joblib\nfrom flask import Flask, request, render_template, redirect, url_for\nfrom modAL.models import ActiveLearner\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.model_selection import train_test_split\n\n# Load the ActiveLearner model\nlearner = joblib.load('active_learner_model.pkl')\n\n# Load the pool data\nX_pool, y_pool = joblib.load('X_y_pool.pkl')\n\n# Split the pool data into train and test sets\nX_pool, X_test, y_pool, y_test = train_test_split(X_pool, y_pool, test_size=0.2, random_state=42)\n\n# Initialize Flask app\napp = Flask(__name__)\n\n# Global variable to store labeled instances\nlabeled_instances = []\n\n@app.route('\/')\ndef home():\n return render_template('index.html')\n\n@app.route('\/query', methods=['GET', 'POST'])\ndef query():\n global X_pool, y_pool, query_instance, query_idx, labeled_instances\n\n if request.method == 'POST':\n # Get the label from the user\n label = int(request.form['label'])\n\n # Teach the model the new label\n learner.teach(X_pool[query_idx].reshape(1, -1), np.array([label]))\n\n # Store the labeled instance and its label\n labeled_instances.append((X_pool[query_idx].reshape(1, -1), label))\n\n # Remove the queried instance from the pool\n X_pool = np.delete(X_pool, query_idx, axis=0)\n y_pool = np.delete(y_pool, query_idx, axis=0)\n\n # Evaluate the model\n accuracy = learner.score(X_test, y_test)\n\n # Query the model for the next instance\n query_idx, query_instance = learner.query(X_pool)\n\n return render_template('query.html', instance=query_instance.tolist(), accuracy=accuracy)\n\n # Initial query\n query_idx, query_instance = learner.query(X_pool)\n accuracy = learner.score(X_test, y_test)\n\n return render_template('query.html', instance=query_instance.tolist(), accuracy=accuracy)\n\n@app.route('\/review')\ndef review():\n accuracy = learner.score(X_test, y_test)\n return render_template('review.html', labeled_instances=labeled_instances, accuracy=accuracy, enumerate=enumerate)\n\n@app.route('\/correct', methods=['POST'])\ndef correct():\n global labeled_instances, learner, X_pool, y_pool\n\n try:\n # Get the index and new label from the form\n index = int(request.form['index'])\n new_label = int(request.form['new_label'])\n\n # Ensure the index is within the valid range\n if 0 <= index < len(labeled_instances):\n # Update the label in the labeled instances\n instance, _ = labeled_instances[index]\n labeled_instances[index] = (instance, new_label)\n\n # Recreate the dataset with corrected labels\n X_corrected = np.vstack([instance for instance, label in labeled_instances])\n y_corrected = np.array([label for instance, label in labeled_instances])\n\n # Clear and reinitialize the learner with corrected data\n learner = ActiveLearner(\n estimator=RandomForestClassifier(),\n X_training=X_corrected,\n y_training=y_corrected\n )\n\n accuracy = learner.score(X_test, y_test)\n\n return render_template('review.html', labeled_instances=labeled_instances, accuracy=accuracy, enumerate=enumerate)\n else:\n return f\"Error: Index {index} is out of range. Valid range is 0 to {len(labeled_instances) - 1}.\", 400\n except Exception as e:\n return str(e), 500\n\nif __name__ == '__main__':\n app.run(debug=True)\n<\/code><\/pre>\n\n\n\nHandling and Correcting Labels<\/h3>\n\n\n\n
review<\/code> route, users can see all labeled instances and correct any mistakes:<\/p>\n\n\n\npythonCopy code
@app.route('\/review')\ndef review():\n accuracy = learner.score(X_test, y_test)\n return render_template('review.html', labeled_instances=labeled_instances, accuracy=accuracy, enumerate=enumerate)\n\n@app.route('\/correct', methods=['POST'])\ndef correct():\n global labeled_instances, learner, X_pool, y_pool\n\n try:\n # Get the index and new label from the form\n index = int(request.form['index'])\n new_label = int(request.form['new_label'])\n\n # Ensure the index is within the valid range\n if 0 <= index < len(labeled_instances):\n # Update the label in the labeled instances\n instance, _ = labeled_instances[index]\n labeled_instances[index] = (instance, new_label)\n\n # Recreate the dataset with corrected labels\n X_corrected = np.vstack([instance for instance, label in labeled_instances])\n y_corrected = np.array([label for instance, label in labeled_instances])\n\n # Clear and reinitialize the learner with corrected data\n learner = ActiveLearner(\n estimator=RandomForestClassifier(),\n X_training=X_corrected,\n y_training=y_corrected\n )\n\n accuracy = learner.score(X_test, y_test)\n\n return render_template('review.html', labeled_instances=labeled_instances, accuracy=accuracy, enumerate=enumerate)\n else:\n return f\"Error: Index {index} is out of range. Valid range is 0 to {len(labeled_instances) - 1}.\", 400\n except Exception as e:\n return str(e), 500\n<\/code><\/pre>\n\n\n\nChallenges and Reflections<\/h3>\n\n\n\n
\n
learner.teach(X_pool[query_idx].reshape(1, -1), np.array([label]))<\/code><\/li>\n\n\n\nX_pool, X_test, y_pool, y_test = train_test_split(X_pool, y_pool, test_size=0.2, random_state=42)<\/code><\/li>\n\n\n\nX_corrected = np.vstack([instance for instance, label in labeled_instances]) y_corrected = np.array([label for instance, label in labeled_instances])<\/code><\/li>\n<\/ol>\n\n\n\n