0% found this document useful (0 votes)

2 views3 pages

Deep Learning Formula Notes

The document outlines the mathematical formulations for Generative Adversarial Networks (GANs) and Long Short-Term Memory (LSTM) networks. It details the functions of the Generator and Discriminator in GANs, including their objective function, and describes the gates and state updates in LSTMs. Additionally, it provides equations for hidden states and outputs in Recurrent Neural Networks (RNNs).

Uploaded by

saurabhhadole3980

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

2 views3 pages

Deep Learning Formula Notes

Uploaded by

saurabhhadole3980

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

GAN Mathematical Formulation

1. Generator Function

G(z) = xfake

Where:

z = Random noise vector

G = Generator network

xfake = Generated fake sample

The Generator transforms random noise into realistic synthetic data.

2. Discriminator Function

D(x) = P(Real)

Where:

x = Input sample

D(x) = Probability that the input is real

D(x) ≈ 1 → Real image

D(x) ≈ 0 → Fake image

3. GAN Objective Function

minG maxD V(D,G) =

Ex~Pdata[log D(x)] +
Ez~Pz[log(1 − D(G(z)))]

Where:

G = Generator

D = Discriminator

Pdata = Distribution of real data

Pz = Distribution of random noise

The Discriminator maximizes the objective by correctly classifying real and fake
samples.
The Generator minimizes it by producing fake samples that fool the Discriminator.

LSTM Mathematical Formulation

1. Forget Gate

ft = σ(Wf[ht−1, xt] + bf)

2. Input Gate

it = σ(Wi[ht−1, xt] + bi)

3. Candidate Memory

C̃t = tanh(Wc[ht−1, xt] + bc)

4. Cell State Update

Ct = ft ⊙ Ct−1 + it ⊙ C̃t

5. Output Gate

ot = σ(Wo[ht−1, xt] + bo)

6. Hidden State

ht = ot ⊙ tanh(Ct)

RNN Equations

Hidden State: ht = tanh(Wxhxt + Whhht−1 + bh)

Output: yt = Whyht + by

Generative Adversarial Networks
No ratings yet
Generative Adversarial Networks
21 pages
GAN Student Notes
No ratings yet
GAN Student Notes
50 pages
Generative Adversarial Network
No ratings yet
Generative Adversarial Network
21 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
30 pages
DL Unit-5
No ratings yet
DL Unit-5
30 pages
Variational Autoencoders and GANs Explained
No ratings yet
Variational Autoencoders and GANs Explained
10 pages
Understanding Generative AI Models
No ratings yet
Understanding Generative AI Models
39 pages
Overview of Generative Adversarial Networks
No ratings yet
Overview of Generative Adversarial Networks
3 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
22 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
19 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
24 pages
Relativistic Discriminator in GANs
No ratings yet
Relativistic Discriminator in GANs
25 pages
Generative Adversarial Networks (GANs)
No ratings yet
Generative Adversarial Networks (GANs)
13 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
21 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
34 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
63 pages
Stock Price Prediction with GANs
No ratings yet
Stock Price Prediction with GANs
8 pages
l11 Post Class
No ratings yet
l11 Post Class
46 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
32 pages
GAN Loss Functions Explained
No ratings yet
GAN Loss Functions Explained
9 pages
Image Generation With GANs - 2
No ratings yet
Image Generation With GANs - 2
111 pages
ML Tech
No ratings yet
ML Tech
8 pages
Overview of GAN Types and Challenges
No ratings yet
Overview of GAN Types and Challenges
8 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
15 pages
GAN Tut
No ratings yet
GAN Tut
17 pages
GAN Architectures and Optimizers Explained
No ratings yet
GAN Architectures and Optimizers Explained
12 pages
Understanding Generative AI Techniques
No ratings yet
Understanding Generative AI Techniques
69 pages
Generative Adversarial Networks (Gans) : Vijay Sankar Data Scientist, Ericsson
No ratings yet
Generative Adversarial Networks (Gans) : Vijay Sankar Data Scientist, Ericsson
21 pages
Introduction to Generative Adversarial Networks
No ratings yet
Introduction to Generative Adversarial Networks
83 pages
GAN Notes
No ratings yet
GAN Notes
30 pages
Overview of Generative Adversarial Networks
No ratings yet
Overview of Generative Adversarial Networks
31 pages
WGAN-GP for Synthetic Loan Data
No ratings yet
WGAN-GP for Synthetic Loan Data
16 pages
Generative Adversarial Network
No ratings yet
Generative Adversarial Network
15 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
14 pages
Generative AI 3
No ratings yet
Generative AI 3
26 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
8 pages
Comprehensive Guide to Generative AI
No ratings yet
Comprehensive Guide to Generative AI
16 pages
Understanding GANs and VAEs in Deep Learning
No ratings yet
Understanding GANs and VAEs in Deep Learning
33 pages
Overview of Generative Adversarial Networks
No ratings yet
Overview of Generative Adversarial Networks
14 pages
DL20 Gan
No ratings yet
DL20 Gan
15 pages
Review of Generative Adversarial Networks
No ratings yet
Review of Generative Adversarial Networks
6 pages
Deep Learning Assignment - V
No ratings yet
Deep Learning Assignment - V
17 pages
Deep Learning Assignment - V
No ratings yet
Deep Learning Assignment - V
17 pages
GAN Session
No ratings yet
GAN Session
68 pages
Overview of Generative Adversarial Networks
No ratings yet
Overview of Generative Adversarial Networks
41 pages
UNIT 4 Notes (Full) .PDF NN
No ratings yet
UNIT 4 Notes (Full) .PDF NN
58 pages
Generative Adversarial Networks (GAN)
No ratings yet
Generative Adversarial Networks (GAN)
64 pages
Deep Generative Modeling with GANs
No ratings yet
Deep Generative Modeling with GANs
27 pages
A Survey On Generative Adversarial Networks (GANs)
No ratings yet
A Survey On Generative Adversarial Networks (GANs)
5 pages
9 2026 01 Deep Learning GAN
No ratings yet
9 2026 01 Deep Learning GAN
45 pages
GANs and ASR Final Exam Guide
No ratings yet
GANs and ASR Final Exam Guide
28 pages
Stock Price Prediction with GANs
No ratings yet
Stock Price Prediction with GANs
15 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
4 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
29 pages
Understanding Generative Models and GANs
No ratings yet
Understanding Generative Models and GANs
19 pages
Understanding Generative Adversarial Networks
No ratings yet
Understanding Generative Adversarial Networks
62 pages
24AML502-DL Unit-5
No ratings yet
24AML502-DL Unit-5
44 pages
WWW - Mtar.in: Parvat Srinivas Reddy
No ratings yet
WWW - Mtar.in: Parvat Srinivas Reddy
20 pages
Explain Adam Optimizer in Detail PDF
No ratings yet
Explain Adam Optimizer in Detail PDF
5 pages
RMSProp Optimizer Note
No ratings yet
RMSProp Optimizer Note
4 pages
Machine Learning Exam Questions Guide
No ratings yet
Machine Learning Exam Questions Guide
2 pages
TCS CodeVita Season 12 Guide
No ratings yet
TCS CodeVita Season 12 Guide
91 pages
Coding and Decoding Practice Questions
No ratings yet
Coding and Decoding Practice Questions
2 pages

Deep Learning Formula Notes

Uploaded by

Deep Learning Formula Notes

Uploaded by

GAN Mathematical Formulation

z = Random noise vector

xfake = Generated fake sample

The Generator transforms random noise into realistic synthetic data.

D(x) = Probability that the input is real

D(x) ≈ 1 → Real image

D(x) ≈ 0 → Fake image

minG maxD V(D,G) =

Pdata = Distribution of real data

Pz = Distribution of random noise

LSTM Mathematical Formulation

ft = σ(Wf[ht−1, xt] + bf)

it = σ(Wi[ht−1, xt] + bi)

C̃t = tanh(Wc[ht−1, xt] + bc)

ot = σ(Wo[ht−1, xt] + bo)

Hidden State: ht = tanh(Wxhxt + Whhht−1 + bh)

You might also like