notes from lecture

ML7331 Module 14 - with mathematical representation

<!-- Google tag (gtag.js) --> <script async src="https://www.googletagmanager.com/gtag/js?id=G-CV2648GQMK"></script> <script> window.dataLayer = window.dataLayer || []; function gtag(){dataLayer.push(arguments);} gtag('js', new Date()); gtag('config', 'G-CV2648GQMK'); </script> Stream Name Texaschikkita Stream URL https://rpubs.com/Texaschikkita Stream ID 9962324179 Measurement Id G-CV2648GQMK

19 November 2024 Capstone A

Comparing Versions

https://www.esri.com/training/Engine/defaultui/player/modern.html?configuration=&preventRightClick=False&cc=en-US&cache=22.16.365.0&playerConfUrl=n&registration=ApiRegistrationId%7C6615b76b8667100028944b12-167021-texaschikkita%21InstanceId%7C0&package=ApiCourseId%7C6615b76b8667100028944b12-167021%21VersionId%7C4&tracking=True&forceReview=False

Feature Class Imaging not working in argis update. Python in arc is trash. I'm gonna try R or use python with API in venv. Or just use QGIS and I'm gonna drop tthe class b c it makes me voilently angry and that's no bueno.

Basic reviews for me for setup on new system - for local use

Comparision and use case for each of shells installed with MSYS2 for C++

gitkraken and sublime merge

compiled in r studio and run in google collab

from asynch canvas, ML documentation, and chatgpt

Follow up lesson to Intro to Python Step 1. From precourse work, bootcamp modules, and course edx classes.

Next lesson in python. Needs some tweaking and cleaning but I'm tired and will do it later.

Intro to python - precourse work and first bootcamp lesson - December 2022. Very basic stuff. Memorize these and practice before moving on. Subsrquent lessons to follow

envs and publishing copied from quarto

Refresher on using Codex (coding Generative PreTrained Transformer). This goes with JBERTquarto doc and RMD on comp. Plan is t test and compare to JBert (jessi's bidirectional encoder repres from transformers) to compare Bert vs Gpt and make better.

from class zoom and notebook: https://colab.research.google.com/drive/1f28mnqRWI2z968JKan5U5AxpycWormen#scrollTo=R7tOqbkMturi

ML 7331 - SMU Fall 2024

POI, Top Restaurants, Accomodations

testing for quarto conversions boring doc

using either mambaforge miniconda or Anaconda. Change commands to conda if using miniconda or Anaconda

ssh spd

working

guide to get expedited passport in dallas

summer project w Larson working outlines

Jessica McPhaul June 9, 2024

April 8, 2024 LAST HW assignment. Stats 2. Spring 2024. Unit 13.

Jessica McPhaul Rafia Mizra Caleb Thornsbury Stas 2 Spring 2024 April 5 2024

April 1, 2024

march 25, 2024

March 25, 2023

March 18, 2024

March 18, 20024

March 11, 2024

March 11, 2024

# Load necessary libraries library(GGally) library(dplyr) # Check if 'censor' column exists in 'training' and 'validate' datasets if (!("censor" %in% names(training)) || !("censor" %in% names(validate))) { # Assuming 'censor' is the last column in the provided subset response_col_name <- names(training)[ncol(training)] # Recode the response variable in both training and validation sets training[[response_col_name]] <- factor(training[[response_col_name]], levels = c(0, 1), labels = c("Malignant", "Benign")) validate[[response_col_name]] <- factor(validate[[response_col_name]], levels = c(0, 1), labels = c("Malignant", "Benign")) } else { # If 'censor' exists, recode directly training$censor <- factor(training$censor, levels = c(0, 1), labels = c("Malignant", "Benign")) validate$censor <- factor(validate$censor, levels = c(0, 1), labels = c("Malignant", "Benign")) } # Generate ggpairs plot for the training data ggpairs(training, columns = 1:9, ggplot2::aes(color = response_col_name))

6372 pre live

Ivan Chavez Rafia Mizra Jessica McPhaul Group Project for class using dataset from Kaggle and WHO

Combine predictions into df ```{r} predictions_df <- data.frame( Actual = actual_values, ANN = predictions_ann, SVM = predictions_svm, RF = predictions_rf, KNN = predictions_knn, LM = predictions_lm ) # Convert to long format for plotting predictions_long <- predictions_df %>% pivot_longer(cols = c("ANN", "SVM", "RF", "KNN", "LM"), names_to = "Model", values_to = "Predicted") ``` Plot together ```{r} ggplot(predictions_long, aes(x = Actual, y = Predicted, colour = Model)) + geom_point() + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "All Models: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy") + theme_minimal() + scale_colour_manual(values = c("blue", "green", "red", "purple", "orange")) ``` ML dataframe ```{r} # Combine all predictions into a single data frame predictions_df <- data.frame( Actual = actual_values, ANN = predictions_ann, SVM = predictions_svm, RF = predictions_rf, KNN = predictions_knn, LM = predictions_lm ) ``` Convert to long for plot ```{r} # Convert to long format for plotting predictions_long <- predictions_df %>% pivot_longer(cols = -Actual, names_to = "Model", values_to = "Predicted") ``` ## Plot all together: ```{r} # Plot all models together ggplot(predictions_long, aes(x = Actual, y = Predicted, colour = Model)) + geom_point() + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "All Models: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy") + theme_minimal() + scale_colour_manual(values = c("blue", "green", "red", "purple", "orange")) ```

# Load necessary libraries library(ggplot2) library(plotly) library(htmlwidgets) # 1. Create the ggplot object with the boxplot gg_boxplot <- ggplot(data, aes(x = Economy_status_Developed, y = Life_expectancy)) + geom_boxplot() + labs(title = "Boxplot of Life Expectancy by Economy Status", x = "Economy Status", y = "Life Expectancy") # 2. Convert the ggplot object to a Plotly object p_plotly <- ggplotly(gg_boxplot) # 3. Display the interactive plot in the RStudio viewer or your web browser # This will show the plot in your RStudio's viewer pane or default web browser print(p_plotly) # 4. Save the Plotly plot as an HTML file htmlwidgets::saveWidget(p_plotly, "Life_Expectancy_by_Status_Plotly.html")

# Load necessary libraries library(ggplot2) library(plotly) library(htmlwidgets) # Create the ggplot object gg_life_expectancy <- ggplot(data, aes(x = Year, y = Life_expectancy, group = Country, color = Country)) + geom_line(alpha = 0.5) + theme_minimal() + theme(legend.position = "none") + labs(title = "Life Expectancy over Time by Country", x = "Year", y = "Life Expectancy") # Convert the ggplot object to a Plotly object p_plotly <- ggplotly(gg_life_expectancy) # Display the interactive plot in the RStudio viewer or your web browser print(p_plotly) # Save the Plotly plot as an HTML file htmlwidgets::saveWidget(p_plotly, "Life_Expectancy_Over_Time_Plotly.html")

# Load necessary libraries library(plotly) library(dplyr) # Select variables select_vars <- c("Life_expectancy", "Adult_mortality", "Alcohol_consumption", "GDP_per_capita", "Schooling") # Create a data frame for pairwise scatter plots pairwise_data <- data %>% select(select_vars) # Create a list to store Plotly scatter plot objects plot_list <- list() # Loop through each combination of variables for pairwise scatter plots for(i in 1:length(select_vars)) { for(j in 1:length(select_vars)) { # Avoid plotting a variable against itself if(i != j) { p <- plot_ly(pairwise_data, x = ~get(select_vars[i]), y = ~get(select_vars[j]), type = 'scatter', mode = 'markers', marker = list(size = 10, opacity = 0.5), xaxis = paste0("x", i), yaxis = paste0("y", j), name = paste(select_vars[i], "vs", select_vars[j])) plot_list[[length(plot_list) + 1]] <- p } } } # Combine all plots final_plot <- subplot(plot_list, nrows = length(select_vars), ncols = length(select_vars), shareX = TRUE, shareY = TRUE) # Display the interactive plot in the RStudio viewer or your web browser print(final_plot) # Save the Plotly plot as an HTML file htmlwidgets::saveWidget(final_plot, "Pairwise_Scatterplots_Plotly.html")

# Plot for Random Forest Model ggplot() + geom_point(aes(x = actual_values, y = predictions_rf), colour = "red") + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "Random Forest Model: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy")

# 5. Plot with box and whisker for RMSE comparison bwplot(results, metric = "RMSE")

# Plot for ANN Model ggplot() + geom_point(aes(x = actual_values, y = predictions_ann), colour = "blue") + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "ANN Model: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy")

# Plot for SVM model ggplot() + geom_point(aes(x = actual_values, y = predictions_svm), colour = "green") + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "SVM Model: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy")

# PLot for KNN Model ggplot() + geom_point(aes(x = actual_values, y = predictions_knn), colour = "purple") + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "KNN Model: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy")

# Plot for LRM ggplot() + geom_point(aes(x = actual_values, y = predictions_lm), colour = "orange") + geom_abline(intercept = 0, slope = 1, linetype = "dashed") + labs(title = "Linear Regression Model: Actual vs Predicted", x = "Actual Life Expectancy", y = "Predicted Life Expectancy")

Analysis from both videos : https://digitalcampus.instructure.com/courses/19956/files/folder/Unit%204?preview=2861314 https://www.youtube.com/watch?v=dQ0hZQUVIqo&list=PPSV file:///C:/Users/Jessica%20M/Desktop/Project%20Presentation%20Video-3.mp4

1-29-24