Skip to content

Assignment 1 Natalia Rodriguez #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
NataRo329 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Assignment 1 Natalia Rodriguez #1

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
54 changes: 42 additions & 12 deletions 02_activities/assignments/assignment_1.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -96,7 +96,7 @@
"metadata": {},
"outputs": [],
"source": [
"# Your answer here"
"wine_df.shape[0]"
]
},
{
Expand All @@ -114,7 +114,7 @@
"metadata": {},
"outputs": [],
"source": [
"# Your answer here"
"wine_df.shape[1]"
]
},
{
Expand All @@ -132,7 +132,8 @@
"metadata": {},
"outputs": [],
"source": [
"# Your answer here"
"wine_df['class'].dtype,\n",
"wine_df['class'].unique()"
]
},
{
Expand All @@ -151,7 +152,7 @@
"metadata": {},
"outputs": [],
"source": [
"# Your answer here"
"wine_df.shape[1]-1"
]
},
{
Expand Down Expand Up @@ -204,7 +205,7 @@
"id": "403ef0bb",
"metadata": {},
"source": [
"> Your answer here..."
"It’s important to standardize predictor variables because it puts them on the same scale, which helps models work better. For example, If one variable ranges from 0–1 and another from 1–10,000, the bigger one can unfairly dominate the model. Also, it is easier to compare effects – In linear models, standardization lets you see which variables matter more. Also, models like logistic regression or neural nets learn faster and more reliably when variables are standardized. Algorithms like KNN or SVM use distances, and without standardization, variables with large scales throw off the results.set.\r\n"
]
},
{
Expand All @@ -220,7 +221,7 @@
"id": "fdee5a15",
"metadata": {},
"source": [
"> Your answer here..."
"The response variable in our case is categorical (e.g., discrete labels like 0, 1, 2), and standardization is only meaningful for continuous numerical data. Standardizing class labels would distort their categorical meaning and interfere with classification algorithms, which rely on clearly defined, separate classes. For example, standardizing labels like 0, 1, and 2 would transform them into continuous values, removing the distinct class boundaries and confusing the model. More generally, even in cases where the response variable is continuous (e.g., blood pressure or probability of disease), standardization is usually avoided unless there's a specific modeling reason. This is because the response variable represents a real-world quantity we want to predict as it exists, and standardizing it would change its scale and interpretation—forcing predictions to be expressed in terms of standard deviations rather than meaningful units."
]
},
{
Expand All @@ -236,7 +237,7 @@
"id": "f0676c21",
"metadata": {},
"source": [
"> Your answer here..."
"It is very important because setting a random seed ensures that random processes are reproducible. Many data science and machine learning tasks involve randomness. Without setting a seed, these processes would yield different results each time you run the code, making replicating results or debugging models hard. The specific seed value is not inherently important — any integer will work. What matters is that the same seed produces the same sequence of random numbers, ensuring consistency and reproducibility. Different seeds will lead to different results, so choosing a fixed seed allows others to replicate your results exactly."
]
},
{
Expand All @@ -261,7 +262,9 @@
"\n",
"# split the data into a training and testing set. hint: use train_test_split !\n",
"\n",
"# Your code here ..."
"response=wine_df['class']\n",
"\n",
"X_train, X_test, y_tarin, y_test=train_test_split(predictors_standardized, response, test_size=0.25, random_state=123)"
]
},
{
Expand Down Expand Up @@ -289,7 +292,20 @@
"metadata": {},
"outputs": [],
"source": [
"# Your code here..."
"#1. Initialize the KNN classifier\n",
"knn = KNeighborsClassifier()\n",
"\n",
"#2. Define the parameter grid: n_neighbors from 1 to 50\n",
"param_grid = {'n_neighbors': list(range(1, 51))}\n",
"\n",
"#3. Set up GridSearchCV with 10-fold cross-validation\n",
"grid_search = GridSearchCV(knn, param_grid, cv=10)\n",
"\n",
"# Fit the model using the training data\n",
"grid_search.fit(X_train, y_train)\n",
"\n",
"#4. Return the best value for n_neighbors\n",
"grid_search.best_params_['n_neighbors'] code here..."
]
},
{
Expand All @@ -310,7 +326,21 @@
"metadata": {},
"outputs": [],
"source": [
"# Your code here..."
"# Retrieve the best k\n",
"best_k = grid_search.best_params_['n_neighbors']\n",
"\n",
"# Initialize the final model with the best number of neighbors\n",
"final_knn = KNeighborsClassifier(n_neighbors=best_k)\n",
"\n",
"# Fit the model on the training data\n",
"final_knn.fit(X_train, y_train)\n",
"\n",
"# Predict on the test set\n",
"y_pred = final_knn.predict(X_test)\n",
"\n",
"# Evaluate accuracy\n",
"accuracy = accuracy_score(y_test, y_pred)\n",
"accuracy"
]
},
{
Expand Down Expand Up @@ -365,7 +395,7 @@
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.10.4",
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
Expand All @@ -379,7 +409,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.19"
"version": "3.12.4"
},
"vscode": {
"interpreter": {
Expand Down