Exploring Super Hero Data and Machine Learning Projects: A Journey of Innovation

Over the past few months, I've been exploring various projects and endeavors, combining the power of machine learning (ML) and artificial intelligence (AI). One of my most recent projects involved an in-depth exploratory data analysis (EDA) on a super hero dataset, which provided a wealth of insights through visualizations using Plotly.

Data Visualization with Plotly

Throughout my analysis, I utilized Plotly to create engaging and informative visualizations. Some of the most interesting plots I generated include:

The most frequently occurring hair colors across DC and Marvel characters The most common characters in both Marvel and DC comic books The most prevalent evil characters in Marvel and DC comics A frequency distribution of the identities of characters in both comic universes The distribution of sexual orientations among characters by gender Homosexual female characters in both Marvel and DC

These plots provide a fascinating glimpse into the characteristics of superheroes and villains across the DC and Marvel universes.

You can view the entire project on GitHub.

Moving Towards Artificial Intelligence and Machine Learning

While I haven't delved extensively into AI yet, my journey in machine learning has been exciting. In recent projects, I've tackled a variety of challenges, from building a line follower robot with my classmate to succeeding in an industry development project aimed at minimizing pollution. I'm particularly proud of a device I created for visually impaired individuals that helps them detect water logs or obstacles in their path while walking.

Machine Learning Code Example

Recently, I worked on a project involving 10 input variables and 1 predictor variable. I developed a function in R that trains multiple classification models to find the one with the least error. Here's a simplified version of the code:

# Function to calculate models
Calculate_models - function(k) {
  j - train[c1:k]
  jFinal - trainExited
  forestModel4 - randomForestFinal ~ ., data  j
  predicted_random3 - predict(forestModel4, validation1[c1:10])
  colnames(predicted_random3) - c("predict")
  merged_pre_act_random3 - merge(predicted_random3, validationExited)
  colnames(merged_pre_act_random3) - c("Predicted", "Actual")
  tfinal2 - table(merged_pre_act_random3)
  confusionMatrix(tfinal2, positive  1)
}
# Function to train models
Train_models - function(to_take) {
  df - combn(1:10, to_take)
  lapply(df, Calculate_models)
}
Train_models(7)

This dataset, the Deep Learning A-Z - ANN dataset, serves as an excellent example of how machine learning techniques can be applied to real-world problems.

Join me on Instagram if you'd like to learn more about my projects and future endeavors.