isaquliyev / ExamPreparationPython

Temporary repository for exam preparations

Exam Preparation Python

1. Scalar, Vector, and Tensor difference:

• Scalar: Singular numerical entity. Examples include loss and accuracy.
• Vector: Ordered collections of numerical values (scalars).
• Tensor: Generalizes the concept of vectors and matrices. Vectors are 1-rank tensors, matrices are 2-rank tensors, and so on.

2. Fashion MNIST and MNIST:

• MNIST: A dataset of handwritten digits.
• Fashion MNIST: A dataset of clothing items.
• Both datasets contain 60,000 training images and 10,000 testing images, with each image being in 28x28 grayscale format. MNIST originally had images in 128x128 format but converted to 28x28.

3. OOP principles:

• Encapsulation, Inheritance, Polymorphism, Abstraction.

4. Python Loops:

• for, while, do while.

5. Compiler vs. Interpreter:

• Compiler translates the whole source code to machine language before the program runs.
• Interpreter reads the code line by line as the code runs.

6. Usage Cases of Sigmoid, Softmax, and ReLU:

• Sigmoid: Binary classification and logistic regression.
• Softmax: Multiclass classification.
• ReLU: Image recognition and computer vision.

7. Python Modules:

• A single Python script containing functions and variables. The module name is the same as the file name without the .py suffix. E.g., file isa.py has the module name isa. Import using import isa.

8. Class vs. Functions:

• Functions are reusable code blocks, while classes are blueprints for creating objects.

9. Data Science and Data Analysis:

• Data analysis involves analyzing past data to inform present decisions. Data science combines data modeling and data collection.

10. DL and ML:

• ML is a subset of AI. DL is a subset of ML. ML uses statistical algorithms and models to improve performance. DL involves neural networks and learns from large datasets.

11. Mutable vs. Immutable:

• Immutable: String, int, tuple, frozenset.
• Mutable: List, dictionary, set. Elements within frozenset and tuple can be mutable.

12. CNN:

• Convolutional Neural Networks are deep learning algorithms used for image classification and object recognition tasks, utilizing three-dimensional data. Example: Face recognition.

13. .py vs. .ipynb:

• .py: Regular Python file containing only code blocks.
• .ipynb: Notebook file containing code blocks, execution results, and internal settings.

14. Usage of Loss Function and Optimizer:

• Loss function shows the difference between predicted data and actual data.
• Optimizer tries to minimize this difference.

15. Tensorflow Operations:

import tensorflow as tf
a = tf.constant([3, 3, 3])
b = tf.constant([2, 2, 2])  # Define tensors

sum_result = tf.add(a, b)    # Addition
diff_result = tf.subtract(a, b)    # Subtraction
quot_result = tf.divide(a, b)    # Division
prod_result = tf.multiply(a, b)    # Multiplication

max, min, abs, log, and exp

16. Set, Tuple, and List:

• Set: A list of unique elements.
• Tuple: Immutable lists.
• List: Mutable lists.

17. Steps in Machine Learning:

• Gathering data
• Pre-processing data (Normalization)
• Training the model
• Optimizing

18. String int bilərsiz də yəqin aq

19. Pandas and Numpy:

• Pandas: Data manipulation and analysis. Missing data shown as NaN.
• Numpy: Used for mathematical operations with arrays, tensors, etc.

20. Shape Operations and Min-Max Operations:

• Shape Operations: Reshaping, Concatenation, Transposing, Splitting, Slicing.
• Min-Max Operations: min, max, argmax, argmin, softmax.