Deep Learning (Neural Networks)

 Deep Learning (Neural Networks): Multi-layered artificial neural networks capable of parsing complex unstructured inputs like images, speech, and text.

Reinforcement Learning

 Reinforcement Learning: Uses an agent that takes actions in an environment and optimizes behavior based on cumulative rewards or penalties rather than static training datasets. 

Semi-Supervised Learning:

 Semi-Supervised Learning: Trains on a small amount of labeled data combined with a large pool of unlabeled data to save labeling costs. 

Unsupervised Learning Models

 Unsupervised Learning Models

Unsupervised models process data without predefined labels, allowing the system to discover structural properties on its own. 

• Clustering: Groups similar data points together based on shared traits (e.g., K-Means, Hierarchical clustering).
• Dimensionality Reduction: Simplifies complex datasets while retaining vital patterns (e.g., Principal Component Analysis). Specialized and Advanced Models

 

Supervised Learning Models

 Supervised Learning Models

Supervised models use input data paired with known output labels to learn the relationship between them.

• Regression models: Predict continuous numeric values like prices or temperatures (e.g., Linear Regression, Ridge, Lasso).
• Classification models: Predict discrete categories or classes (e.g., Logistic Regression, Support Vector Machines, Naive Bayes).
• Decision Trees & Ensembles: Combine multiple decision rules or trees to boost accuracy (e.g., Random Forest).

 

 Different type of models used in Machine       Learning in Python 

Scikit-Learn (Traditional Machine Learning)

LinearRegression: Basic regression for continuous data.
LogisticRegression: Standard baseline for binary classification.
RandomForestClassifier / RandomForestRegressor: Powerful ensemble tree models.
GradientBoostingClassifier: Sequential tree building for high accuracy.
SVC: Support Vector Classifier for complex boundaries.
KMeans: Unsupervised clustering algorithm.

XGBoost & LightGBM (Gradient Boosting)

XGBClassifier / XGBRegressor: High-performance, scalable gradient boosting.
LGBMClassifier / LGBMRegressor: Fast, leaf-wise tree growth models.

PyTorch (Deep Learning Ecosystem)

torchvision.models.resnet50: Industry standard for image classification.
torchvision.models.vit_b_16: Vision Transformer for advanced image tasks.
torchaudio.models.wave2vec2_model: Architecture for speech processing.
torch.nn.Transformer: Raw building block for sequence-to-sequence tasks.

TensorFlow / Keras (Deep Learning Ecosystem)

keras.applications.ResNet50: Pre-trained deep residual network for vision.
keras.applications.MobileNetV3Large: Lightweight, mobile-optimized vision model.
keras.applications.EfficientNetB0: State-of-the-art scaling for image tasks.
keras.layers.LSTM: Recurrent layer stringed together for text/time-series.

Hugging Face Transformers (NLP & GenAI)

bert-base-uncased: Baseline for text classification and extraction.
roberta-base: Improved, harder-trained version of BERT.
gpt2: Standard starting point for causal text generation.
meta-llama/Meta-Llama-3-8B: State-of-the-art large language model for fine-tuning.
google/vit-base-patch16-224: Vision transformer adapted for Hugging Face pipelines. 

Statsmodels (Statistical & Time Series)

OLS: Ordinary Least Squares regression with deep statistical summaries.
ARIMA: Autoregressive Integrated Moving Average for time-series forecasting.
Logit: Logistic regression specialized for statistical inference. 

HOW TO TRAIN A MODEL STEP BY STEP IN PYTHON

 

HOW TO TRAIN A MODEL STEP BY STEP IN PYTHON:

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

from sklearn.model_selection import train_test_split

from sklearn.linear_model import LinearRegression

from sklearn.metrics import mean_squared_error, r2_score

data = {

'Hours': [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],

'Scores': [10, 20, 30, 40, 50, 60, 70, 80, 85, 95]

}

df = pd.DataFrame(data)

print(df.head())

   Hours  Scores
0      1      10
1      2      20
2      3      30
3      4      40
4      5      50




plt.scatter(df['Hours'], df['Scores'])

plt.xlabel('Hours Studied')

plt.ylabel('Scores')

plt.title('Hours vs Scores')

plt.show()





X = df[['Hours']]  # Independent variable

y = df['Scores']   # Dependent variable

#Now split into training and testing sets:

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)


model = LinearRegression()

model.fit(X_train, y_train)

y_pred = model.predict(X_test)

print(y_pred)

[88.01724138 20.73275862]
print("Mean Squared Error:", mean_squared_error(y_test, y_pred))

print("R2 Score:", r2_score(y_test, y_pred))
Mean Squared Error: 4.820340368608807
R2 Score: 0.9954363641480627
y_pred = model.predict(X_test)

print(y_pred)


[88.01724138 20.73275862]

print("Mean Squared Error:", mean_squared_error(y_test, y_pred))

print("R2 Score:", r2_score(y_test, y_pred))

Mean Squared Error: 4.820340368608807
R2 Score: 0.9954363641480627


plt.scatter(X, y, color='blue')

plt.plot(X, model.predict(X), color='red')

plt.xlabel('Hours Studied')

plt.ylabel('Scores')

plt.title('Linear Regression - Hours vs Scores')

plt.show()




hours = np.array([[7.5]])

predicted_score = model.predict(hours)

print(f"Predicted Score for 7.5 hours: {predicted_score[0]}")
Predicted Score for 7.5 hours: 73.59913793103448