From 14d8ed99e64b1115c9a92b6776837bfed86a2494 Mon Sep 17 00:00:00 2001
From: mbentle6 <68927153+mbentle6@users.noreply.github.com>
Date: Mon, 5 Oct 2020 18:44:02 -0700
Subject: [PATCH] Update HW_4.ipynb

Fixed typo in Problem 5, fixed language in Problem 6 and added the "Z" variable for clarification
---
 lesson_4_Training/HW_4.ipynb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/lesson_4_Training/HW_4.ipynb b/lesson_4_Training/HW_4.ipynb
index 420b1ce..79d5a86 100644
--- a/lesson_4_Training/HW_4.ipynb
+++ b/lesson_4_Training/HW_4.ipynb
@@ -198,7 +198,7 @@
     "id": "QhSnTaNuzOd4"
    },
    "source": [
-    "Create create Y, X, and Z samples from problem 1 with 100,000,000 observations (100 million). Measure how long does it take to estimate linear regression of Y on X and Z using: Stochastic Gradient Descent, Batch Gradient Descent, Linear Regression from sklearn, and Linear regression estimated using matrix multiplication $$\\hat{\\theta} = (\\pmb{X}^T \\cdot \\pmb{X})^{-1} \\cdot \\pmb{X}^T \\cdot \\pmb{y}$$, and Linear regression estimated preudo-inverse. For Batch Gradient Descent use 1000 iterations and eta = 0.01. For Stochastic Gradient Descent use 5 iterations and eta = 0.01. Use SGDRegressior imported from sklearn.model.\n",
+    "Create Y, X, and Z samples from problem 1 with 100,000,000 observations (100 million). Measure how long does it take to estimate linear regression of Y on X and Z using: Stochastic Gradient Descent, Batch Gradient Descent, Linear Regression from sklearn, and Linear regression estimated using matrix multiplication $$\\hat{\\theta} = (\\pmb{X}^T \\cdot \\pmb{X})^{-1} \\cdot \\pmb{X}^T \\cdot \\pmb{y}$$, and Linear regression estimated preudo-inverse. For Batch Gradient Descent use 1000 iterations and eta = 0.01. For Stochastic Gradient Descent use 5 iterations and eta = 0.01. Use SGDRegressior imported from sklearn.model.\n",
     "\n",
     "For each estimation import **MSE** and **time**"
    ]
@@ -288,7 +288,7 @@
    },
    "source": [
     "# Problem 6\n",
-    "Recreate a X,Y data from problem 1 using 500 observations. Create a 25-degree polynomial for X and scale the data using standard scaler. Estimate the regression model predicting Y via Ridge regression. Calculate MSE (average 'neg_mean_squared_error') using cross-validation with cv = 3. Find optimal alpha by looping from 0.0001 to 1 using step size 0.0001."
+    "Recreate the X,Y and Z data from problem 1 using 500 observations. Create a 25-degree polynomial for X and scale the data using standard scaler. Estimate the regression model predicting Y via Ridge regression. Calculate MSE (average 'neg_mean_squared_error') using cross-validation with cv = 3. Find optimal alpha by looping from 0.0001 to 1 using step size 0.0001."
    ]
   },
   {