Skip to content

Latest commit

 

History

History
47 lines (43 loc) · 1.35 KB

DAY32.md

File metadata and controls

47 lines (43 loc) · 1.35 KB

DAY 32

Nosql

  1. Document Oriented Database --> MongoDB
  2. Column Oriented Database ---> Google Big Table,Dynamo DB,Hbase,Cassendra
  3. Key:Value type Nosql -----.> Redis
  4. Graph DB ------> Neo4j

MongoDB

AWS Lambda (Lambda + DYnamoDB)

  • lambda---> code

Tensorflow

  • tensor ---> dimensons

NN (Neural Networks)

use cases

  • performance Search
  • Voice processing
  • Self Driving Car
  • Computer Games

Artificial Neuaral Network (ANN)

  • input in Neuron ---> Dendrite
  • process and store---->cell body
  • output of Neuron----> Axon
  • Neuron ------> in NN Nodes
  • Input layer--->Hidden Layer---> Output layer (complex problems)
  • input layer----> Output layer (simple problems)
  • in input layer attribute = Neuron
  • each neuron in input layer connect with hidden layer every neuron(depends on us)
  • jitni hidden layer utne CPU,GPU,TPU required
  • if we us More than two hidden layer its called DEEP NEURAL NETWORK

Cost Function

  • check the error in actual output using cost function

Activation FUNCTION

  • its Process input equation in hidden layer

Types

  1. Threshold
  2. Sigmoid function
  3. Rectifier function
  4. Hyperbolic Tangent

Back Prapogation

  • Q--> Comparision b/w outputs
  • wait adjust responsibility --> gradient descent
  1. Batch Gradient Descent
  2. Stochastic Gradient Descent

Epoc