Deep Learning - Function Approximation

中文筆記
Kaggle

Problem Description

The dataset contains the data which are the outputs from an unknown function $y=f(x_1,x_2)$ Note that the given training dataset has been disturbed by some noise. You are now required to design a network to remove the introduced noise. After finishing your network training, you need to produce the outputs for a given testing dataset. Kaggle will compute the mean square error (MSE) between your outputs and the true outputs for the performance evaluation of your network.

Problem Evaluation

The criterion used to evaluate your submission is the MSE loss $L=\frac{1}{n}\sum_{i=1}^n(\hat{y}_i-y)^2$

Dataset Description

Two datasets are given:

• train.csv: the training dataset which contains three columns of data. Column 1 and Column 2 are inputs x1 and x2, while Column 3 is the target value y.
• test.csv: the testing dataset for evaluating your model. Only two columns of data, i.e., x1 and x2, are given.