Jun Hyuk Kim's Blog

Web crawling and web scraping are miss understood. I had some problems because the only term I knew was wab crawling. Web crawling is used to find URL's from a web page where we can web scrape or use the URL to get indexes that can be used to create a search index. You can think of this as a spider trying to create a web connecting URLs together. First a starting URL is used to get all the URL's..

Convolutional Neural Network is mostly used in image data. Because the data is in a grid a discreate convolution is used. The filter is moved over the data and the overlapping cells are multiplied. Then the multiplied values are added to get the value for the data. This filter can be used for data other than images. Because the filter has its dimensions the CNN can be used in sequential data to ..

Momentum Use the moving average to move the data. Due to having the previous momentum as a value it reduces large changes. Think of a ball rolling down a hill into a flat surface. Creates cases where the ball moves in a different direction from the slope. Nesterov accelerated gradient Uses the gradient from the expected direction. Fixes the problem of moving in a different direction from the min..

Stochastic gradient descent is when the model uses one data at a time to update the parameters. This is great at being efficient in terms of memory. Creates noisy gradients due to it being only one data. Can be used to escape the local minimum. Batch Gradient Descent Batch gradient descent is when the model uses the all the data at once to update the parameters. This must keep all the values to ..