UNIT 5 CIE

Edit Answers
123456789101112131415161718192021222324252627
Across
  1. 3. Devices like smartphones and IoT units that can run AI locally.
  2. 8. Settings (like learning rate) that control how an ML model learns.
  3. 10. In Matplotlib, it defines the color range used in a plot.
  4. 13. Process of analyzing datasets visually before model training (short form).
  5. 14. Rule-of-thumb algorithms often used in AI problem-solving.
  6. 15. Type of Seaborn chart to compare categorical data values.
  7. 16. Hardware accelerators from Google used to train large AI models efficiently.
  8. 19. Tool that uses genetic programming to automate ML pipeline creation.
  9. 21. Algorithm like SGD or Adam that adjusts weights to minimize loss.
  10. 22. TensorFlow mechanism to control training behavior (e.g., early stopping).
  11. 24. One full pass of training data through the ML algorithm.
  12. 25. Python tool for automated machine learning (AutoML).
  13. 26. Library for numerical operations, often used with Matplotlib and Pandas.
  14. 27. Python Imaging Library used in image processing tasks.
Down
  1. 1. Common metric used to evaluate classification models.
  2. 2. Web-based notebook used for interactive Python coding, often in AI.
  3. 4. Tool used to normalize data before feeding into machine learning models.
  4. 5. Python library for classical machine learning tasks like classification/regression.
  5. 6. AI chatbot from Google, often compared with ChatGPT.
  6. 7. Core data structure used in Pandas for storing tabular data.
  7. 9. When a model performs well on training data but poorly on new data.
  8. 10. Module alias used when importing OpenCV in Python.
  9. 11. Software that is freely available and modifiable, like TensorFlow and PyTorch.
  10. 12. General term for representing data graphically using tools like Matplotlib.
  11. 15. Type of error in AI models when predictions are systematically off.
  12. 17. When a model is too simple to capture the underlying patterns.
  13. 18. Vector that guides neural networks in updating weights via backpropagation.
  14. 20. A single unit of input (word/piece of a word) processed by language models.
  15. 23. Function that measures how well the model's prediction matches the actual output.