Neural Networks – Data Business Intelligence

1. Feedforward Neural Network (FNN)

Overview:

A basic type of neural network where information moves in one direction, from input to output, through hidden layers.

Use Cases:

Social Media: Predict user engagement based on post attributes (time, content type, hashtags).
Cybersecurity: Classify network traffic as normal or anomalous based on network features.
SEO Analysis: Predict website traffic based on SEO metrics (keyword density, backlinks, page speed).

Code Example (Python, Keras):

from keras.models import Sequential

from keras.layers import Dense
import numpy as np

# Sample Data
X = np.random.rand(100, 10)  # 100 samples, 10 features
y = np.random.randint(2, size=(100, 1))  # Binary output

# Build Feedforward Neural Network
model = Sequential()
model.add(Dense(64, input_dim=10, activation='relu'))
model.add(Dense(32, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

# Compile and train the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.fit(X, y, epochs=10, batch_size=10)

2. Convolutional Neural Network (CNN)

Overview:

CNNs are designed for processing structured grid data like images, but can also be used for text analysis and time-series data.

Use Cases:

Social Media: Image classification for user-generated content (e.g., categorizing images as memes, ads, or personal posts).
Cybersecurity: Detect anomalies in image-based network patterns (e.g., heatmaps of traffic).
SEO Analysis: Analyze website screenshots to assess visual SEO elements (layout, images, etc.).

Code Example (Python, Keras):

from keras.models import Sequential

from keras.layers import Conv2D, MaxPooling2D, Flatten, Dense
import numpy as np

# Sample Data: Image-like input
X = np.random.rand(100, 64, 64, 3)  # 100 samples of 64x64 RGB images
y = np.random.randint(2, size=(100, 1))

# Build CNN
model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(64, 64, 3)))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Flatten())
model.add(Dense(64, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

# Compile and train the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.fit(X, y, epochs=10, batch_size=10)

3. Recurrent Neural Network (RNN)

Overview:

RNNs are designed to handle sequential data. They maintain hidden states that allow them to remember previous inputs, making them ideal for time-series analysis.

Use Cases:

Social Media: Predict future user engagement based on historical post interactions.
Cybersecurity: Detect anomalous sequences of network events.
SEO Analysis: Predict traffic trends based on historical keyword rankings.

Code Example (Python, Keras):

from keras.models import Sequential

from keras.layers import SimpleRNN, Dense
import numpy as np

# Sample Data: Sequential data
X = np.random.rand(100, 10, 1)  # 100 samples, 10 time steps, 1 feature per step
y = np.random.randint(2, size=(100, 1))

# Build RNN
model = Sequential()
model.add(SimpleRNN(50, activation='relu', input_shape=(10, 1)))
model.add(Dense(1, activation='sigmoid'))

# Compile and train the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.fit(X, y, epochs=10, batch_size=10)

4. Long Short-Term Memory (LSTM)

Overview:

LSTM is a type of RNN that can capture long-term dependencies in sequential data. It’s particularly effective when the relevant information may be far apart in the sequence.

Use Cases:

Social Media: Predict content virality by modeling user behavior over time.
Cybersecurity: Detect advanced persistent threats (APTs) by analyzing long sequences of user activity or traffic logs.
SEO Analysis: Predict long-term changes in keyword rankings based on past data trends.

Code Example (Python, Keras):

from keras.models import Sequential

from keras.layers import LSTM, Dense
import numpy as np

# Sample Data: Sequential data
X = np.random.rand(100, 10, 1)  # 100 samples, 10 time steps, 1 feature per step
y = np.random.randint(2, size=(100, 1))

# Build LSTM model
model = Sequential()
model.add(LSTM(50, activation='relu', input_shape=(10, 1)))
model.add(Dense(1, activation='sigmoid'))

# Compile and train the model
model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
model.fit(X, y, epochs=10, batch_size=10)

5. Autoencoder

Overview:

Autoencoders are unsupervised neural networks used for dimensionality reduction and anomaly detection. They learn to compress data into a lower-dimensional space and then reconstruct it.

Use Cases:

Social Media: Detect anomalies in user behavior, such as unusual posting patterns.
Cybersecurity: Identify anomalous traffic patterns by detecting deviations in reconstructed data.
SEO Analysis: Detect unusual website behavior, such as sudden drops in traffic or ranking changes.

Code Example (Python, Keras):

from keras.models import Model

from keras.layers import Input, Dense
import numpy as np

# Sample Data
X = np.random.rand(100, 10)  # 100 samples, 10 features

# Build Autoencoder
input_layer = Input(shape=(10,))
encoded = Dense(5, activation='relu')(input_layer)
decoded = Dense(10, activation='sigmoid')(encoded)

autoencoder = Model(input_layer, decoded)

# Compile and train the model
autoencoder.compile(optimizer='adam', loss='mse')
autoencoder.fit(X, X, epochs=10, batch_size=10)

Summary:

Feedforward Neural Network (FNN): Simple predictions (user engagement, traffic classification).
Convolutional Neural Network (CNN): Image-based analysis (user-generated content, website screenshots).
Recurrent Neural Network (RNN): Sequential data modeling (engagement or attack trends).
Long Short-Term Memory (LSTM): Long-term dependencies (predicting virality or APT detection).
Autoencoder: Anomaly detection (unusual patterns in social media, SEO metrics, or network traffic).

References
[1] https://wiki.pathmind.com/neural-network
[2] https://tirendazacademy.medium.com/artificial-neural-networks-in-machine-learning-fa653d74b1a1