-
Book Overview & Buying
-
Table Of Contents
Machine Learning For Dummies
By :
Finding the right coefficients for a linear model is just a matter of time and memory. However, sometimes a system won’t have enough memory to store a huge dataset. In this case, you must resort to other means, such as learning from one example at a time rather than having all of them loaded into memory. The following sections help you understand the one-example-at-a-time approach to learning.
Using gradient descent
The gradient descent finds the right way to minimize the cost function one iteration at a time. After each step, it accounts for all the model’s summed errors and updates the coefficients in order to make the error even smaller during the next data iteration. The efficiency of this approach derives from considering all the examples in the sample. The drawback of this approach is that you must load all the data into memory.
Unfortunately, you can’t always store all the data in memory because some datasets are huge. In addition...
Change the font size
Change margin width
Change background colour