TensorFlow Fundamentals

What is TensorFlow?

Open-source ML library for:

• Data preprocessing
• Model building
• Model serving

Key Concepts

Tensors

Multi-dimensional numerical representations:

• Scalar: 0D (single number)
• Vector: 1D array
• Matrix: 2D array
• Tensor: nD array

Creating Tensors

import tensorflow as tf

# Constants
scalar = tf.constant(7)
vector = tf.constant([10, 10])
matrix = tf.constant([[10, 7], [7, 10]])

# Variables (mutable)
var = tf.Variable([10, 7])

Tensor Operations

• Addition, subtraction, multiplication
• Matrix multiplication: tf.matmul()
• Aggregations: min, max, mean, sum
• Reshaping: tf.reshape(), tf.transpose()

GPU Acceleration

# Check GPU availability
tf.config.list_physical_devices('GPU')

# TensorFlow automatically uses GPU when available

Best Practices

• Use tf.function for speed
• Leverage GPU/TPU when possible
• Start simple, then optimize
• Profile and monitor performance

Master tensors to build powerful ML models with TensorFlow.