Now that home security systems are readily available at a low cost, these systems are commonly being installed to watch over homes and loved ones. These systems are fairly easy to install and can provide 4k Ultra HD resolution. The user can configure the sensitivity and areas to monitor and receive object detection notifications. Unfortunately, once the customer starts to use the system, they often find that the notifications are overwhelming and soon turn them off. After hearing the same experience from multiple friends and family I thought it would be a good topic for my thesis. I examined a top selling security system sold at a bulk retail store and have implemented improved detection techniques that advance object detection and reduce false notifications. The additional algorithms will support the processing of both near real-time streams and saved video file processing, which existing security systems do not include.

Low-level optimization is the process of handwriting key parts of applications in assembly code that is better than what can be generated from a higher-level language. In performance-intensive applications, this is key to ensuring efficient code. This is generally something that is taught in on the job training, but knowledge of it improves college student’s skill sets and makes them more desirable employees I have created material for a course teaching this low-level optimization with assembly code. I specifically focus on the x86 architecture, as this is one of the most prolific computer architectures. The course contains a series of lecture videos, live coding videos, and structured programming assignments to support the learning objectives. This material is presented in an entirely autonomous way, which serves as remote learning material and can be easily added as supplemental material to an existing course.

Machine learning is the process of training a computer with algorithms to learn from data and make informed predictions. In a world where large amounts of data are constantly collected, machine learning is an important tool to analyze this data to find patterns and learn useful information from it. Machine learning applications expand to numerous fields; however, I chose to focus on machine learning with a business perspective for this thesis, specifically e-commerce.

The e-commerce market utilizes information to target customers and drive business. More and more online services have become available, allowing consumers to make purchases and interact with an online system. For example, Amazon is one of the largest Internet-based retail companies. As people shop through this website, Amazon gathers huge amounts of data on its customers from personal information to shopping history to viewing history. After purchasing a product, the customer may leave reviews and give a rating based on their experience. Performing analytics on all of this data can provide insights into making more informed business and marketing decisions that can lead to business growth and also improve the customer experience.
For this thesis, I have trained binary classification models on a publicly available product review dataset from Amazon to predict whether a review has a positive or negative sentiment. The sentiment analysis process includes analyzing and encoding the human language, then extracting the sentiment from the resulting values. In the business world, sentiment analysis provides value by revealing insights into customer opinions and their behaviors. In this thesis, I will explain how to perform a sentiment analysis and analyze several different machine learning models. The algorithms for which I compared the results are KNN, Logistic Regression, Decision Trees, Random Forest, Naïve Bayes, Linear Support Vector Machines, and Support Vector Machines with an RBF kernel.

Behavioral economics suggests that emotions can affect an individual’s decision making. Recent research on this idea’s application on large societies hints that there may exist some correlation or maybe even some causation relationship between public sentiment—at least what can be pulled from Twitter—and the movement of the stock market. One major result of consistent research on whether or not public sentiment can predict the movement of the stock market is that public sentiment, as a feature, is becoming more and more valid as a variable for stock-market-based machine learning models. While raw values typically serve as invaluable points of data, when training a model, many choose to “engineer” new features for their models—deriving rates of change or range values to improve model accuracy.
Since it doesn’t hurt to attempt to utilize feature extracted values to improve a model (if things don’t work out, one can always use their original features), the question may arise: how could the results of feature extraction on values such as sentiment affect a model’s ability to predict the movement of the stock market? This paper attempts to shine some light on to what the answer could be by deriving TextBlob sentiment values from Twitter data, and using Granger Causality Tests and logistic and linear regression to test if there exist a correlation or causation between the stock market and features extracted from public sentiment.

The goal of this product was to create a highly customizable application in which any individual, musician or not, can create a harmony for the user’s melody. This Automating Music Composer is built on the underlying rules of music composition, rules that are unique for each type of music available. This program is built on rules that are similar to how a Finite State Machine works (Fig 1). Each state represents a different chord in a given key, where the first roman numeral represents the first note in the chord progression. Each transition represents the action that can be taken by the chord progression, or the next note that can be reached by the current note. The user is able to manipulate these rules and styles, adjust different musical parameters to their liking, and is able to input their own melody, which then will output a unique harmony. This product aims to bridge the gap between predictive technologies and musical composition. Allowing the user to be more involved in the composition process helps the program to act as a tool for the user, rather than a separate entity that simply gives the user a completed recording. This allows the user to appreciate and understand what they are helping to produce more than they would if they were to simply be an inactive consumer of a random music composer. This product is meant to feel like an extension of the user, rather than a separate tool.

The Tutoring Center Management System is a web-based application for ASU’s University Academic Success Programs (UASP) department, particularly the Math Tutoring Center. It is aimed at providing a user-friendly interface to track queue requests from students visiting the tutoring centers and convert that information into actionable data with the potential to live-track and assess the performance of each tutoring center and each tutor. Numerous UASP processes are streamlined to create an efficient and integrated workflow, such as tutor scheduling, tutor search, shift coverage requests, and analytics. The intended users of the application feature ASU students and the UASP staff, including tutors and supervisors.

In this update to the ESPBot, we have introduced new libraries for a small OLED display and a beeper. This functionality can be easily expanded to multiple beepers and displays, but requires more GPIO pins, or for the user to not use some of the infrared sensors or the ultrasonic sensor. We have also relocated some of the pins. The display can be updated to display 1 of 4 predefined shapes, or to display user-defined text. New shapes can be added by defining new methods within display.ino and calling the appropriate functions while parsing the JSON data in viple.ino. The beeper can be controlled by user-defined input to play any frequency for any amount of time. There is also a function added to play the happy birthday song. More songs can be added by defining new methods within beeper.ino and calling the appropriate functions while parsing the JSON data in viple.ino. More functionality can be added to allow the user to input a list of frequencies along with a list of time so the user can define their own songs or sequences on the fly.

Machine learning is one of the fastest growing fields and it has applications in almost any industry. Predicting sports games is an obvious use case for machine learning, data is relatively easy to collect, generally complete data is available, and outcomes are easily measurable. Predicting the outcomes of sports events may also be easily profitable, predictions can be taken to a sportsbook and wagered on. A successful prediction model could easily turn a profit. The goal of this project was to build a model using machine learning to predict the outcomes of NBA games.
In order to train the model, data was collected from the NBA statistics website. The model was trained on games dating from the 2010 NBA season through the 2017 NBA season. Three separate models were built, predicting the winner, predicting the total points, and finally predicting the margin of victory for a team. These models learned on 80 percent of the data and validated on the other 20 percent. These models were trained for 40 epochs with a batch size of 15.
The model for predicting the winner achieved an accuracy of 65.61 percent, just slightly below the accuracy of other experts in the field of predicting the NBA. The model for predicting total points performed decently as well, it could beat Las Vegas’ prediction 50.04 percent of the time. The model for predicting margin of victory also did well, it beat Las Vegas 50.58 percent of the time.

This thesis dives into the world of machine learning by attempting to create an application that will accurately predict whether or not a sneaker will resell at a profit. To begin this study, I first researched different machine learning algorithms to determine which would be best for this project. After ultimately deciding on using an artificial neural network, I then moved on to collecting data, using StockX and Twitter. StockX is a platform where individuals can post and resell shoes, while also providing statistics and analytics about each pair of shoes. I used StockX to retrieve data about the actual shoe, which involved retrieving data for the network feature variables: gender, brand, and retail price. Additionally, I also retrieved the data for the average deadstock price for each shoe, which describes what the mean price of new, unworn shoes are selling for on StockX. This data was used with the retail price data to determine whether or not a shoe has been, on average, selling for a profit. I used Twitter’s API to retrieve links to different shoes on StockX along with retrieving the number of favorites and retweets each of those links had. These metrics were used to account for ‘hype’ of the shoe, with shoes traditionally being more profitable the larger the hype surrounding them. After preprocessing the data, I trained the model using a randomized 80% of the data. On average, the model had about a 65-70% accuracy range when tested with the remaining 20% of the data. Once the model was optimized, I saved it and uploaded it to a web application that took in user input for the five feature variables, tested the datapoint using the model, and outputted the confidence in whether or not the shoe would generate a profit.
From a technical perspective, I used Python for the whole project, while also using HTML/CSS for the front-end of the application. As for key packages, I used Keras, an open source neural network library to build the model; data preprocessing was done using sklearn’s various subpackages. All charts and graphs were done using data visualization libraries matplotlib and seaborn. These charts provided insight as to what the final dataset looked like. They showed how the brand distribution is relatively close to what it should be, while the gender distribution was heavily skewed. Future work on this project would involve expanding the dataset, automating the entirety of the data retrieval process, and finally deploying the project on the cloud for users everywhere to use the application.