![\"Open](\"https://colab.research.google.com/assets/colab-badge.svg\" "\\"Open"){kind=icon}
![\"Download\"](\"https://img.shields.io/badge/Github-Download-blue.svg\" "\\"Download"){kind=icon}
"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "wkbLGIXiXDLA",
"nbpages": {
"level": 1,
"link": "[5.1 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1-Ridge-Regression)",
"section": "5.1 Ridge Regression"
}
},
"source": [
"# 5.1 Ridge Regression"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "im3dFZNuXDLI",
"nbpages": {
"level": 1,
"link": "[5.1 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1-Ridge-Regression)",
"section": "5.1 Ridge Regression"
}
},
"source": [
"Created by Ben Whewell (bwhewell@nd.edu)"
]
},
{
"cell_type": "markdown",
"metadata": {
"nbpages": {
"level": 1,
"link": "[5.1 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1-Ridge-Regression)",
"section": "5.1 Ridge Regression"
}
},
"source": [
"This example was adapted from:\n",
"\n",
"McClarren, Ryan G (2018). Uncertainty Quantification and Predictive Computational Science: A Foundation for Physical Scientists and Engineers, Chapter 4: Local Sensitivity Analysis Based on Derivative Approximations, Springer, https://doi.org/10.1007/978-3-319-99525-0_4"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "YCh0H5_wXDLM",
"nbpages": {
"level": 1,
"link": "[5.1 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1-Ridge-Regression)",
"section": "5.1 Ridge Regression"
}
},
"outputs": [],
"source": [
"## import all needed Python libraries here\n",
"import numpy as np\n",
"from sklearn.datasets import load_boston\n",
"import matplotlib.pyplot as plt\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.linear_model import RidgeCV"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "lGFZ3g_eXSJt",
"nbpages": {
"level": 1,
"link": "[5.1 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1-Ridge-Regression)",
"section": "5.1 Ridge Regression"
}
},
"outputs": [],
"source": [
"# Functions to be used Later\n",
"np.random.seed(47)\n",
"# Train/Test Split\n",
"def tt_split(x,y,size):\n",
" inds = np.arange(len(x))\n",
" np.random.shuffle(inds) \n",
" split = int(np.ceil(len(x)*size))\n",
" # x_train, x_test, y_train, y_test\n",
" return x[inds[split:]],x[inds[:split]],y[inds[split:]],y[inds[:split]]\n",
"\n",
"# Mean Squared Error\n",
"def mse(y,y_hat):\n",
" return np.mean((y-y_hat)**2)"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "oJKTEmM1XrJj",
"nbpages": {
"level": 2,
"link": "[5.1.1 Sensitivities with Least-Squares Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.1-Sensitivities-with-Least-Squares-Regression)",
"section": "5.1.1 Sensitivities with Least-Squares Regression"
}
},
"source": [
"## 5.1.1 Sensitivities with Least-Squares Regression\n",
"QoI $Q(x)$ where $\\mathbf{x} = (x_1, ..., x_J)^T$ of $J$ parameters. \n",
"$I$ Equations that relate the known values of $Q_i$ and $\\mathbf{x}_i$ to the unknown sensitivities.\n",
"\n",
"$X_{ij} = (x_{ij} - \\bar{x}_j) \\quad$ \n",
"$\\mathbf{y} = \\begin{pmatrix} Q_1 - Q(\\mathbf{\\bar{x}}) \\\\ Q_2 - Q(\\mathbf{\\bar{x}})\\\\ ... \\\\ Q_I - Q(\\mathbf{\\bar{x}}) \\end{pmatrix} \\quad $ $\\beta = \\begin{pmatrix} \\left.\\frac{\\partial Q}{\\partial x_1}\\right|_\\mathbf{\\bar{x}} \\\\ \\left.\\frac{\\partial Q}{\\partial x_2}\\right|_\\mathbf{\\bar{x}} \\\\ ... \\\\ \\left.\\frac{\\partial Q}{\\partial x_J}\\right|_\\mathbf{\\bar{x}} \\end{pmatrix}$\n",
"\n",
"This can be rearranged into $\\mathbf{X\\beta} = \\mathbf{y}$, which is similar to a Least Squares problem when $I > J$ and where $\\hat{\\beta}_{LS} = \\left(\\mathbf{X}^T\\mathbf{X} \\right)^{-1}\\mathbf{X}^T\\mathbf{y}$.\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "jns-MNPKbqAi",
"nbpages": {
"level": 2,
"link": "[5.1.2 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.2-Ridge-Regression)",
"section": "5.1.2 Ridge Regression"
}
},
"source": [
"## 5.1.2 Ridge Regression\n",
"#### Normalize the Data\n",
"\n",
"$X_{ij} = \\frac{(x_{ij} - \\bar{x}_j)}{\\bar{x}_j} \\quad$ \n",
"$\\beta = \\begin{pmatrix} \\left.\\bar{x}_1\\frac{\\partial Q}{\\partial x_1}\\right|_\\mathbf{\\bar{x}} \\\\ \\left.\\bar{x}_2\\frac{\\partial Q}{\\partial x_2}\\right|_\\mathbf{\\bar{x}} \\\\ ... \\\\ \\left.\\bar{x}_J\\frac{\\partial Q}{\\partial x_J}\\right|_\\mathbf{\\bar{x}} \\end{pmatrix}$\n",
"\n",
"Ridge regression adds the $\\ell_2$ penalty to the $\\beta$ minimization problem. This tries to minimize the resulting $\\beta$ vector.\n",
"\\begin{equation}\n",
"\\hat{\\beta}_{\\text{ridge}} = \\min_{\\beta} \\sum_{i =1}^{I} (y_i - \\beta \\cdot \\mathbf{x}_i)^2 + \\lambda \\beta||_2 \\quad \\\\ ||\\beta||_2 = \\left(\\sum_{i=1}^{I}|\\beta_i|^2\\right)^{1/2} \n",
"\\end{equation}\n",
"\n",
"\\begin{equation}\n",
"\\hat{\\beta}_{\\text{ridge}} = \\left(\\mathbf{X}^T\\mathbf{X} +\\lambda \\mathbf{I} \\right)^{-1}\\mathbf{X}^T\\mathbf{y}\n",
"\\end{equation}\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "4Ovq_zuSXeJ9",
"nbpages": {
"level": 2,
"link": "[5.1.2 Ridge Regression](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.2-Ridge-Regression)",
"section": "5.1.2 Ridge Regression"
}
},
"outputs": [],
"source": [
"# Ridge Regression\n",
"def ridge_model(x_train,y_train,Lambda,x_test,coef_=False):\n",
" beta = np.linalg.inv(x_train.T @ x_train + Lambda *np.eye(x_train.shape[1])) @ x_train.T @ y_train\n",
" if coef_:\n",
" return beta\n",
" return x_test @ beta"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "hVi-ygDMd2_X",
"nbpages": {
"level": 2,
"link": "[5.1.3 Cross Validation](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.3-Cross-Validation)",
"section": "5.1.3 Cross Validation"
}
},
"source": [
"## 5.1.3 Cross Validation\n",
"$\\lambda$ is a hyperparameter that needs to be optimized through cross validation. \n",
"\n",
"#### Process for CV\n",
"1. Select number of folds ($K$-folds)\n",
"2. Select $\\lambda = (\\lambda_1, ... ,\\lambda_L)$ values\n",
"3. $\\texttt{For k in folds:}$ \\\\\n",
" $\\qquad \\texttt{Randomly sample data leaving $1/k$ of data as testing}$ \\\\\n",
" $ \\qquad \\texttt{for jj in $\\lambda$:}$ \\\\\n",
" $ \\qquad \\qquad \\texttt{Calculate $\\hat{y}$ of Ridge Regression} $ \\\\\n",
" $ \\qquad \\qquad \\texttt{Calculate MSE between $\\hat{y}$ and $y$} $\n",
"4. A matrix of size $k$-folds $\\times$ $\\lambda$ averages the $\\lambda$ across the folds\n",
"5. The largest $\\lambda$ value is chosen that is still within one standard error of the smallest mean MSE.\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "jEC1-FRPeB4_",
"nbpages": {
"level": 2,
"link": "[5.1.3 Cross Validation](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.3-Cross-Validation)",
"section": "5.1.3 Cross Validation"
}
},
"outputs": [],
"source": [
"# Cross Validation\n",
"def cross_validation(x,y,Lambda,k_folds):\n",
" # Error matrix\n",
" error = np.zeros((k_folds,len(Lambda)))\n",
" for ii in range(k_folds):\n",
" # Resample the training and testing data\n",
" x_train,x_test,y_train,y_test = tt_split(x,y,1/k_folds)\n",
" for jj in range(len(Lambda)):\n",
" # Fit Model to Different Lambda values\n",
" y_hat = ridge_model(x_train,y_train,Lambda[jj],x_test)\n",
" # Calculate MSE \n",
" error[ii,jj] = mse(y_test,y_hat)\n",
" return best_lambda(error,Lambda)\n",
" \n",
"# Calculating Optimal lambda for Ridge Regression\n",
"def best_lambda(error,Lambda):\n",
" # Maximum lambda within 1 standard error of minimum of mean MSE\n",
" bound = np.min(np.mean(error,axis=0))+np.std(error)/np.sqrt(k_folds*len(Lambda)) \n",
" means = np.mean(error,axis=0)\n",
" best = Lambda[0] # initialize \n",
" for ii in range(len(Lambda)):\n",
" if (bound - means[ii]) > 0 and Lambda[ii] > best:\n",
" best = Lambda[ii]\n",
" # return Lambda[np.argmin(means)] # minimum MSE\n",
" return best\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "tmNSYKVIjXze",
"nbpages": {
"level": 2,
"link": "[5.1.4 Example Problem](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.4-Example-Problem)",
"section": "5.1.4 Example Problem"
}
},
"source": [
"## 5.1.4 Example Problem\n",
"#### I > J "
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "Jo8K9xb5jWid",
"nbpages": {
"level": 2,
"link": "[5.1.4 Example Problem](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.4-Example-Problem)",
"section": "5.1.4 Example Problem"
}
},
"outputs": [],
"source": [
"np.random.seed(47)\n",
"# I > J (Least Squares Problem, I = 506, J = 13)\n",
"X,Y = load_boston(return_X_y=True)\n",
"# Normalize Data - Dimensionless\n",
"Y = Y - np.mean(Y)\n",
"X = (X - np.mean(X,axis=0))/np.mean(X,axis=0)\n",
"# Search different lambdas\n",
"Lambdas = [0.01,0.1,0.15,0.2,0.5,1,10,20]\n",
"k_folds = len(X) # Leave one out validation\n",
"# Calculating Optimal Lambda\n",
"param_ = cross_validation(X,Y,Lambdas,k_folds=k_folds)\n",
"# Train/Test split (20% Testing)\n",
"x_train,x_test,y_train,y_test = tt_split(X,Y,0.2)\n",
"# Calculate coefficients in Ridge Regression\n",
"beta_1 = ridge_model(x_train,y_train,param_,x_test,coef_=True)"
]
},
{
"cell_type": "markdown",
"metadata": {
"colab_type": "text",
"id": "s6jwdTIVjlX1",
"nbpages": {
"level": 4,
"link": "[5.1.4.1 I < J ](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.4.1-I-<-J)",
"section": "5.1.4.1 I < J "
}
},
"source": [
"#### 5.1.4.1 I < J "
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"colab": {},
"colab_type": "code",
"id": "6CPcK3lMjxcF",
"nbpages": {
"level": 4,
"link": "[5.1.4.1 I < J ](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.4.1-I-<-J)",
"section": "5.1.4.1 I < J "
}
},
"outputs": [],
"source": [
"np.random.seed(47)\n",
"# I < J (Estimating Local Sensitivities, I = 10, J = 13)\n",
"X,Y = load_boston(return_X_y=True)\n",
"# Randomly Chose 10 simulations\n",
"inds = np.random.choice(range(0,len(X)),10)\n",
"x = X[inds]; y = Y[inds]\n",
"# Normalize Data - Dimensionless\n",
"y = y-np.mean(y)\n",
"x = (x - np.mean(x,axis=0))/np.mean(x,axis=0)\n",
"# Search different lambdas\n",
"Lambdas = [0.01,0.1,0.15,0.2,0.5,1,10,20]\n",
"k_folds = len(x) # Leave one out validation\n",
"# Calculating Optimal Lambda\n",
"param_ = cross_validation(X,Y,Lambdas,k_folds=k_folds)\n",
"# Train/Test split (20% Testing)\n",
"x_train,x_test,y_train,y_test = tt_split(x,y,0.2)\n",
"# Calculate coefficients in Ridge Regression\n",
"beta_2 = ridge_model(x_train,y_train,param_,x_test,coef_=True)"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 511
},
"colab_type": "code",
"id": "FeDy20S9jyP1",
"nbpages": {
"level": 4,
"link": "[5.1.4.1 I < J ](https://ndcbe.github.io/cbe67701-uncertainty-quantification/05.01-Contributed-Example.html#5.1.4.1-I-<-J)",
"section": "5.1.4.1 I < J "
},
"outputId": "67ed148e-a88f-4b8b-e7b3-c7cc168ae7b4"
},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
""
]
}
],
"metadata": {
"colab": {
"collapsed_sections": [],
"name": "05.01-Contributed-Example.ipynb",
"provenance": []
},
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.3"
}
},
"nbformat": 4,
"nbformat_minor": 1
}