A Linear Regression with Non Linear Basis Functions

[joshthekid]

So I am currently learning some regression techniques for my research and have been reading a text that describes linear regression in terms of basis functions. I got linear basis functions down and no exactly how to get there because I saw this a lot in my undergrad basically, in matrix notation
y=w^Tx
you then define your loss function as
1/n Σⁿ(w_i*x_i-y_i)²
then you take the partial derivatives with respect to w set it equal to zero and solve.

So now I want to use a non-linear basis functions, let's say I want to use m gaussians basis functions, φ_i, the procedure is the same but I am not sure exactly on the construction of the model. Let's say I have L features is the model equation of the form

y_n=Σ^mΣ^Lw_iφ_i(x_j)

in other words I have created a linear combination of M new features, φ(x), which are constructed with all L of the previous features for each data point n:
y_n=w₀+w₁(φ₁(x₁)+φ₁(x₂)...+...φ₁(x_L) ...+...w_m(φ₁(x₁)+φ₂(x₂)...+...φ_m(x_L))

where x_i are features / variables for my model and not data values? I hope this makes sense. Thanks in advance.

 

[micromass]

The parameters you wish to estimate are the $w_i$ and the values $(x_1,...,x_L)$ are known for each data point?

 

[joshthekid]

The parameters you wish to estimate are the $w_i$ and the values $(x_1,...,x_L)$ are known for each data point?

That is correct.

 

[micromass]

Then you have a standard linear regression. Linear refers to the coefficients and not the functions used. Thus your loss function is again

L = \sum_{i=1}^n \left(y_i - w_0 - w_1\sum_k \phi_1(x_k) - w_2\sum_k \phi_2(x_k) - ... - w_N \sum_k \phi_N(x_k)\right)^2

and you minimize this by taking partial derivatives and setting them equal to $0$. In matrix notation, you let $Y$ by the column matrix with entries the $y_i$ and you let $X$ be the design matrix whose $i$th row is
\left(1~~\sum_k \phi_1(x_k)~~ ...~~\sum_k \phi_N(x_k)\right)
The coefficients are then $W = (X^TX)^{-1} X^T Y$.

 

[joshthekid]

Then you have a standard linear regression. Linear refers to the coefficients and not the functions used. Thus your loss function is again

L = \sum_{i=1}^n \left(y_i - w_0 - w_1\sum_k \phi_1(x_k) - w_2\sum_k \phi_2(x_k) - ... - w_N \sum_k \phi_N(x_k)\right)^2

and you minimize this by taking partial derivatives and setting them equal to $0$. In matrix notation, you let $Y$ by the column matrix with entries the $y_i$ and you let $X$ be the design matrix whose $i$th row is
\left(1~~\sum_k \phi_1(x_k)~~ ...~~\sum_k \phi_N(x_k)\right)
The coefficients are then $W = (X^TX)^{-1} X^T Y$.

Great Thanks, this is what I thought it meant but the way you wrote it makes it lot clearer than the text I am using which has all formulas in matrix notation and it hard to tell if they are talking about a single random variable or a vector of random variables.