Baidu Artificial Intelligence study notes L1 concept

1. Artificial intelligence and machine learning, deep learning of the relationship:
   
   machine learning, especially in supervised learning, there are more specific to refer to. Machine learning is specializing in computer simulation or how to achieve human learning behavior to acquire new knowledge or skills, re-organize existing knowledge structures so as to continuously improve their performance.
   The step of machine learning may be realized in two steps, training, and prediction. The two professional terms similar to the meaning of induction and deduction. Induction is an abstract general rule, machine learning "training" is also true from a specific case. From certain number of samples (known model input and model output Y X), the learning of the input-output relationship X Y (the expression can be thought of some kind). It is derived from the interpretation of the specific cases of the general rule results, machine learning prediction is also true.
   The relationship between Y and X training obtained based on the input face of the emerging X calculated output Y. Most of the time, the output of the model calculated, and if the same output in the real scene, that the model is effective.
   H is assumed in the model, which is a parameter regarding $\theta$ functions and input X of a function $H(\theta, X)$ represents. Objective optimization model is to make $H(\theta, X)$ output and real value $Y$ as consistent as possible. So, the process of learning parameters is known in the sample, decreasing the evaluation function ( $H(\theta, X)$ and $Y$ difference) of the process until a parameter learned $\theta$ such that the minimum value of the evaluation function. This model predictive value and measure the gap between the true value of the evaluation function is also called loss function (loss Loss). Optimization of the parameters of the above process as shown in FIG formula: