Python is a dominant language in artificial intelligence (AI) and machine learning (ML), supported by a robust ecosystem of libraries. This overview covers key libraries that form the foundation for AI and data science workflows.

NumPy

NumPy is a fundamental package for numerical computing in Python. It provides a efficient multidimensional array object, ndarray, along with a comprehensive collection of functions for fast array operations. NumPy's capabilities in linear algebra, random number generation, and Fourier transforms make it indispensable for scientific computing, serving as the base for many higher-level libraries.

Code Example: Generate a NumPy array and compute the sum of its elements.

import numpy as np

sample_data = np.array([5, 10, 15, 20])
print("Array:", sample_data)
print("Sum:", np.sum(sample_data))

Pandas

Pandas is a powerful data manipulation and analysis library built on NumPy. It introduces two primary data structures: Series for one-dimensional labeled data and DataFrame for two-dimensional, tabular data. Pandas excels at data cleaning, transformation, aggregation, and visualization, making it a go-to tool for data preparation in AI pipelines.

Code Example: Construct a DataFrame from a list of dictionaries and filter rows.

import pandas as pd

records = [
    {"Product": "Laptop", "Price": 1200, "Stock": 15},
    {"Product": "Mouse", "Price": 25, "Stock": 80},
    {"Product": "Keyboard", "Price": 75, "Stock": 45}
]

df = pd.DataFrame(records)
print(df)
filtered = df[df["Stock"] > 50]
print("High stock items:\n", filtered)

Matplotlib

Matplotlib is a comprehensive plotting library for creating static, animated, and interactive visualizations in Python. It offers a MATLAB-like interface for generating a wide variety of plots, including line charts, bar graphs, scatter plots, and histograms. Its flexibility and control over plot elements make it essential for data exploration and result presentation.

Code Example: Create a bar chart to compare values.

import matplotlib.pyplot as plt

categories = ["A", "B", "C", "D"]
values = [12, 19, 8, 15]

plt.bar(categories, values, color="skyblue")
plt.title("Value Comparison")
plt.xlabel("Category")
plt.ylabel("Value")
plt.show()

Scikit-learn

Scikit-learn is a user-friendly machine learning library that provides efficient implementations of a broad range of algorithms for classification, regression, clustering, and dimensionality reduction. It includes tools for model selection, evaluation, preprocessing, and data splitting, following a consistent API that simplifies the ML workflow from prototyping to evaluation.

Code Example: Train a k-nearest neighbors classifier on synthetic data.

from sklearn.neighbors import KNeighborsClassifier
from sklearn.datasets import make_classification
from sklearn.model_selection import train_test_split

X, y = make_classification(n_samples=200, n_features=4, random_state=42)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)

knn = KNeighborsClassifier(n_neighbors=3)
knn.fit(X_train, y_train)
score = knn.score(X_test, y_test)
print("Model accuracy:", score)

TensorFlow

TensorFlow is an open-source framework for building and deploying machine learning models, with a focus on deep learning. Its core abstraction is the computational graph, where operation are represented as nodes and data flows as tensors. The high-level Keras API, integrated into TensorFlow, simplifies the creation of neural networks, enabling rapid development of models for tasks like image recognition and natural language processing.

Code Example: Build and compile a convolutional neural network for image classification.

import tensorflow as tf

cnn_model = tf.keras.Sequential([
    tf.keras.layers.Conv2D(32, (3, 3), activation="relu", input_shape=(64, 64, 3)),
    tf.keras.layers.MaxPooling2D((2, 2)),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(64, activation="relu"),
    tf.keras.layers.Dense(10, activation="softmax")
])

cnn_model.compile(optimizer="adam",
                  loss="categorical_crossentropy",
                  metrics=["accuracy"])
cnn_model.summary()