Boostcamp AI Tech 3 : Computer Vision(CV)

Let's commit any code everyday.

Machine_Learning_Loadmap.jpg

Week 1 keyword : AI Math & Python

• AI Math

  • Norm
  • Inner Product
  • Array
  • Linear Regression with inverse array
  • Gradient Descent(GD)
  • Linear regression with GD
  • mini batch Stochastic Gradient Descent(mini batch SGD)
  • Deep Learning
  • Monte Carlo Sampling
  • Maximum Likelihood Estimator(MLE)
  • Bayesian Theory
  • Convolutional Neural Network(CNN)
  • Recurrent Neural Network(RNN)

• Python

  • Outline
    • Dynamic Typing
    • Object-Oriented Programming(OOP)
    • Interpreter
  • Variables
    • Memory
  • Function and Console IO
    • Function
    • Console in/out
    • Formatting
  • Conditionals and Loops
    • Condition
    • Loop
    • Debug
  • String and advanced function concept
    • String
    • Call by Object Reference
    • Function - scoping rule
    • Recursive Function
    • Docstring
    • PEP8
  • Python data structure
    • Stack
    • Queue
    • Tuple
    • Set
    • Dict
    • collections
      • deque
      • OrderedDict
      • defaultdict
      • Counter
      • namedtuple
  • Pythonic code
    • split & join
    • list comprehension
    • enumerate & zip
    • lambda & map & reduce
    • generator
    • function passing arguments
      • keyward arguments
      • default arguments
      • variable-length asterisk(*args)
      • keyward-length asterisk(**kwargs)
      • asterisk(unpacking a container)
  • Object-Oriented Programming(OOP)
    • outline
    • objects in python
    • OOP charactoeristics
      • inheritance
      • polymorphism
      • visibility
    • decorate
      • first-class objects
      • inner function
      • decorator
  • Module and Project
    • module
    • package
    • virtual environment

Week 1 memoir

Week 2 keyword : PyTorch

• Pytorch
  • Operation
    • Tensor
    • view vs reshape
    • squeeze, unsqueeze
    • dot, mm, matmul
    • nn.functional
    • AutoGrad
  • torch.nn.Module
    • nn.Parameter
    • Backward
    • Dataset
    • DataLoader
    • model.save()
    • checkpoints
  • Monitoring Tools
    • Tensorboard
    • weight & biases
  • Multi-GPU
    • Model parallel
    • Data parallel
      • DataParallel
      • DistributedDataParallel
  • Hyperparameter Tuning
  • Ray
  • Troubleshooting

Week 2 memoir

Week 3 keyword : Data Visualization

• Iris
• StudentsPerformance
• prices
• heart

Week 3 memoir

Week 4 keyword : Deep Learning Basic

• Deep Learning Basic
• Optimization
• CNN(Convolutional Neural Network)
• RNN(Recurrent Neural Network)
• Generative Model

Week 4 memoir

Week 5 keyword : AI Service Development Basic

• Machine Lreaning Project Life Cycle
• Linux & Shell Command
• Docker
• MLflow

Week 5 memoir

Week 6,7 keyword : Classification

• P Stage Start!
• Data Feeding
• Model
• Training & Inference
• More...
• Why image data need to normalization?

Week 6 memoir

Week 7 memoir

Week 8 keyword : Classification, Semantic Segmemtation, Object Detection, CNN Visualization

Directory

Week 8 memoir