Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Introduction to Deep Learning and Neural Networks.
Search
Bedanta Bikash Borah
July 24, 2018
Education
0
260
Introduction to Deep Learning and Neural Networks.
Bedanta Bikash Borah
July 24, 2018
Tweet
Share
More Decks by Bedanta Bikash Borah
See All by Bedanta Bikash Borah
Thinking beyond platforms with KMP
iambedant
0
9
Let's Stream that Video - an ExoPlayer Starters Guide
iambedant
0
130
A tale of Multiplatform
iambedant
0
91
A Full-Stack app with Kotlin —by an Android Developer
iambedant
2
230
Other Decks in Education
See All in Education
Tutorial: Foundations of Blind Source Separation and Its Advances in Spatial Self-Supervised Learning
yoshipon
1
150
Transición del Management al Neuromanagement
jvpcubias
0
210
DIP_1_Introduction
hachama
0
120
JOAI2025講評 / joai2025-review
upura
0
420
仮説の取扱説明書/User_Guide_to_a_Hypothesis
florets1
4
380
令和政経義塾第2期説明会
nxji
0
200
Data Management and Analytics Specialisation
signer
PRO
0
1.5k
OpenSourceSummitJapanを運営してみた話
kujiraitakahiro
0
790
生態系ウォーズ - ルールブック
yui_itoshima
1
250
AI for Learning
fonylew
0
180
フィードバックの伝え方、受け身のココロ / The Way of Feedback: Words and the Receiving Heart
spring_aki
1
140
日本の情報系社会人院生のリアル -JAIST 修士編-
yurikomium
1
120
Featured
See All Featured
Build your cross-platform service in a week with App Engine
jlugia
231
18k
Performance Is Good for Brains [We Love Speed 2024]
tammyeverts
12
1.1k
The Straight Up "How To Draw Better" Workshop
denniskardys
236
140k
Bash Introduction
62gerente
615
210k
Code Review Best Practice
trishagee
70
19k
GraphQLの誤解/rethinking-graphql
sonatard
72
11k
A designer walks into a library…
pauljervisheath
207
24k
Building Applications with DynamoDB
mza
96
6.6k
Testing 201, or: Great Expectations
jmmastey
45
7.7k
How to Create Impact in a Changing Tech Landscape [PerfNow 2023]
tammyeverts
53
2.9k
Automating Front-end Workflow
addyosmani
1370
200k
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
252
21k
Transcript
Introduction to Deep Learning and Neural Networks. Bedanta Bikash Borah
@iamBedant
None
None
Agenda 1. What is Deep Learning? 2. Why Deep Learning
is taking off? 3. How Deep Learning works? 4. Training. 5. Example MNIST. 6. Code Sample. 7. Few extra concepts.
Deep Learning The term Deep Learning refers to training very
large Neural Network
What is Neural Network? House Size (X) Price (Y) 500
5 Laks 600 8 Laks 700 9 Laks 800 13 Laks 900 12 Laks 1100 18 Laks 1200 20 Laks
Housing Price Prediction 0 laks 5 laks 10 laks 15
laks 20 laks 000 sq ft 300 sq ft 600 sq ft 900 sq ft 1200 sq ft What is Neural Network?
What is Neural Network? x Y Size of the House
Price of the House Neuron Function F Input Output
What is Neural Network? Size No of bedrooms Locality Society
Y Family Size Walkability School/Market Quality
What is Neural Network? X1 X2 X3 X4 Y Size
No of bedrooms Locality Society
Deep Neural Network
Why DeepLearning taking off? Deep Neural Network Medium Neural Network
Traditional ML Algorithm Data Performance Not well defined
Why DeepLearning taking off? 1. Data 2. Computation 3. Algorithm
How DeepLearning Works? X1 X2 X3 X4 Y
How DeepLearning Works? X1 X2 X3 X4 Y f W1
W2 W3 W4 X1* W1 + X2 * W2 + X3 * W3 +X4 * W4 Z f( )= Relu (x) or Sigmoid(x) *ignoring bias for simplification Z =
How DeepLearning Works?
How DeepLearning Works? X1 X2 X3 X4 Y
Interviewer: What is your biggest strength? Me: I am an
expert in machine learning. Interviewer: What’s 9 + 10? Me: It’s 3. Interviewer: Nowhere near. It’s 19. Me: It’s 16. Interviewer: Wrong. It’s still 19 Me: It’s 18. Interviewer: No. It’s 19 Me: It’s 19. Interviewer: You’re hired
Training X1 X2 X3 X4 Y Y’ Random Initialisation
Training 1.Quadratic cost 2.Cross-entropy cost 3.Exponential cost Cost Function:
Training Grad Gradient Descent
Training Learning Rate (alpha)
Training Large Learning rate
Training Learning Rate (alpha)
Training New Weights = Existing Weights Learning Rate - *
Gradient ( )
Interviewer: What is your biggest strength? Me: I am an
expert in machine learning. Interviewer: What’s 9 + 10? Me: It’s 3. Interviewer: Nowhere near. It’s 19. Me: It’s 16. Interviewer: Wrong. It’s still 19 Me: It’s 18. Interviewer: No. It’s 19 Me: It’s 19. Interviewer: You’re hired.
Example MNIST
MNIST 60,000 training samples 10,000 test samples
MNIST
MNIST = 28 x 28 = 784
MNIST Softmax Image Vector Neural Network Layers
MNIST Softmax Image Vector Neural Network Layers
MNIST Softmax Image Vector Neural Network Layers
MNIST Softmax Image Vector Neural Network Layers
MNIST Softmax Image Vector Neural Network Layers
MNIST **from three blue one brown’s “But, what is a
neural network?” video**
Talk is cheap show me the code.
Advanced MNIST CNN (convolutional neural network)
Interviewer: What is your biggest strength? Me: I am an
expert in machine learning. Interviewer: What’s 9 + 10? Me: It’s 3. Interviewer: Nowhere near. It’s 19. Me: It’s 16. Interviewer: Wrong. It’s still 19 Me: It’s 18. Interviewer: No. It’s 19 Me: It’s 19. Interviewer: You’re hired.
Interviewer: What is your biggest strength? Me: I am an
expert in machine learning. Interviewer: What’s 9 + 10? Me: It’s 3. Interviewer: Nowhere near. It’s 19. Me: It’s 16. Interviewer: Wrong. It’s still 19 Me: It’s 18. Interviewer: No. It’s 19 Me: It’s 19. Interviewer: What’s 20 + 10? Me: It’s 19
Advanced MNIST Overfitting (Regularization, Dropout)
Extras
Reference https://github.com/iamBedant/CMRIT-Deeplearning-TechTalk-Demo Simple MNIST Example https://github.com/iamBedant/TensoreFlowLite Android TFLite Example Others:
https://www.tensorflow.org/ https://keras.io/
Thank You !!! @iamBedant