Harshil Patel's repositories
Iot-Cyber-Security-with-Machine-Learning-Research-Project
IoT networks have become an increasingly valuable target of malicious attacks due to the increased amount of valuable user data they contain. In response, network intrusion detection systems have been developed to detect suspicious network activity. UNSW-NB15 is an IoT-based network traffic data set with different categories for normal activities and malicious attack behaviors. UNSW-NB15 botnet datasets with IoT sensors' data are used to obtain results that show that the proposed features have the potential characteristics of identifying and classifying normal and malicious activity. Role of ML algorithms is for developing a network forensic system based on network flow identifiers and features that can track suspicious activities of botnets is possible. The ML model metrics using the UNSW-NB15 dataset revealed that ML techniques with flow identifiers can effectively and efficiently detect botnets’ attacks and their tracks.
IoT-Network-Intrusion-Detection-and-Classification-using-Explainable-XAI-Machine-Learning
The continuing increase of Internet of Things (IoT) based networks have increased the need for Computer networks intrusion detection systems (IDSs). Over the last few years, IDSs for IoT networks have been increasing reliant on machine learning (ML) techniques, algorithms, and models as traditional cybersecurity approaches become less viable for IoT. IDSs that have developed and implemented using machine learning approaches are effective, and accurate in detecting networks attacks with high-performance capabilities. However, the acceptability and trust of these systems may have been hindered due to many of the ML implementations being ‘black boxes’ where human interpretability, transparency, explainability, and logic in prediction outputs is significantly unavailable. The UNSW-NB15 is an IoT-based network traffic data set with classifying normal activities and malicious attack behaviors. Using this dataset, three ML classifiers: Decision Trees, Multi-Layer Perceptrons, and XGBoost, were trained. The ML classifiers and corresponding algorithm for developing a network forensic system based on network flow identifiers and features that can track suspicious activities of botnets proved to be very high-performing based on model performance accuracies. Thereafter, established Explainable AI (XAI) techniques using Scikit-Learn, LIME, ELI5, and SHAP libraries allowed for visualizations of the decision-making frameworks for the three classifiers to increase explainability in classification prediction. The results determined XAI is both feasible and viable as cybersecurity experts and professionals have much to gain with the implementation of traditional ML systems paired with Explainable AI (XAI) techniques.
dynamic-programming-model-MATLAB-for-environmental-investment-decision-making-in-coal-mining
Coal is the widespread fossil fuel on earth. It provides the necessary material foundation for economic development of a country. However, coal mining activities cause a lot of environmental impacts that are hazardous to the health of citizens in mining regions and place costs on the government. According to government laws and regulations, coal mines should invest in related pollution treatment projects to meet the emission standards. How to allocate the limited resources among a set of pollutant treatment projects to minimize the total losses, including penal loss and vacancy loss, from an investment perspective is a typical decision-making problem. Therefore, the present study proposed a discrete dynamic programming procedure to provide an effective solution for decision-making in treatment project investment. Furthermore, a case study involving the Laojuntang coal mine of Zhengzhou Coal Industry (Group) of China on the treatment project investment problem was implemented using the proposed model. The results demonstrate that the proposed model is effective and applicable for environmental investment decision-making at a typical coal mine in terms of minimizing the total losses.
ML-Models-Comparion-Indian-Liver-Data-Set
Comparison of ML Models with ILPD (Indian Liver Patient Dataset) Data Set which contains 10 variables that are age, gender, total Bilirubin, direct Bilirubin, total proteins, albumin, A/G ratio, SGPT, SGOT and Alkphos.
Nonlinear-Optimization-on-fed-batch-reactor-system
This survey relates to using nonlinear optimization to improve or optimize fed-batch reactor system regarding the discussion in the academic journal presented “Optimization of fed-batch bioreactors using genetic algorithm: multiple control variables” by Debasis Sarkar and Jayant M. Modak, published in Computers & Chemical Engineering. Furthermore, the original article, “Optimal Production of Secreted Protein in Fed-Batch Reactors” by Seujeung Park and W. Fred Ramirez, published in AIChE Journal, Volume 34. Additionally, it should be acknowledged this report was produced for ISE 3210 (Nonlinear and Dynamic Optimization) at Ohio State University and is not meant for reproduction or publication.
Deep-Learning-with-Glass-Identification-Data-Set
The dataset is provided by University of California at Irvine Machine Learning Repository. For this example, we use the glass dataset created by B. German.The purpose of this dataset is classifying different types of glass based on the level on certain components, such as Potassium, Aluminium, etc
Discrete-Event-Simulation-of-Agricultural-Inventory-Managment-System-Project
The Inventory Management System is a system that buys ingredients, stores excess ingredients, and creates cattle and chicken feed. The three ingredients that create cattle and chicken feed are corn, lime, and fish meal. There are subtle differences in the makeup of cattle and chicken feed, however all three products are used in both feeds. The system is an s-S inventory management model. This means that when a specific product is below a threshold, an order for that product will be processed. The suppliers have an order sheet detailing the cost of different size ingredient purchasing options and if the option has the ability to be shipped overnight or must be shipped in the standard 3-5 days time. The raw material costs are copied in the planning section, see Table 2. If the Inventory Management System has left over ingredients then they must be stored. Raw ingredients are stored in a manner that allows for one thousand pounds of ingredients to be stored in a twelve square foot block. The payments for storage space are one dollar and eight cents per square foot with an additional cost of fourteen to twenty cents per square foot in utility/overhead cost. The system does not have defined inventory evaluation conditions or feed prices and is, firstly, in need of assessment. After assessment, the company must improve the system to maximize profits. This is where the team can provide incredible insight via simulation. In order to assess performance, the easiest process is to experiment with the actual system - however this Inventory Management system current has no success settings. This would mean that all experiments would be blind and could only be evaluated after enough data is accrued. This would be very detrimental to the business. If an experiment were to go awry, customers would no longer buy from the company and the business would take a very large financial hit. Since experimenting with the actual system is not an option, the next step would be to experiment with a mathematical model. Since there are multiple undefined variables, the process of analytically defining a solution would be very tedious. Finding an analytical model might prove successful with one set of variables and price points, but would not lend itself to the ability to pivot on variable values without having to recalculate final solutions. The best route to finding optimal settings for the Inventory Management System is simulation. Simulation has the ability to pivot on variable values and quickly recalculate immediate impact and impact on future orders. In addition, the ability to simulate a “regular” week based on historic distributions as well as the ability to simulate light or heavy order weeks would provide immense value to the company. In order to provide a concrete solution, the team will be designing an inventory system to maximize profit for the system. The goal is to set two prices, P1 and P2, for the price of cattle and chicken feed, respectively, that would routinely net the company profit. The price cannot exceed a markup of 45% from the breakeven price for a pound of feed (calculated from the per pound cost from Table 2). This will eliminate the ability to charge an unrealistic amount for the product to offset a poorly planned business model. The solution will be achieved by defining when the company checks inventory, conditions for purchasing more inventory, price to charge per pound of feed, process of handling back-orders, and the monthly square feet of storage
Long-short-term-memory-AI-LSTM-on-Time-Series-Airline-Passengers-Dataset-Data
This is a very short exploration into applying LSTM techniques using the Keras library. Code and content is based on several cool blog posts and papers; see references section for more info.
Markov-Decision-Process-Toolbox-Practice_Simple
A Markov decision process (MDP) is a discrete time stochastic control process. It provides a mathematical framework for modeling decision making in situations where outcomes are partly random and partly under the control of a decision maker.
ML-Regression-Methods-Comparison-with-Wine-Quality-Dataset
ML Regression/Prediction Model Comparison using Wine Quality Data Set from UCI. The goal is to model wine quality based on physicochemical tests
Address-Finder-by-Name
Find a Address with Name of location
AI-SNAKE-GAME-PROJECT-IN-PYTHON-FIRST-TIME-USING-COLAB
The goal of the snake fame is to eat as many apples as possible while avoiding your tail from the AI operated snake
AirBNB-Hacker
Python Script to Find Last-minute Deals on AirBNB by setting up notifications on new deals
Decision-Trees-Classifer_Iris-Data
Using the iris dataset to illustrate Decision Tree and Neural Nets. This is a common practice dataset that contains the sepal and petal length and width of 150 iris flowers of three different species: Iris-Setosa, Iris-Versicolor, and Iris-Virginica
HealthCare-Spending-and-GDP-TO-Evaluate-Life-Expenctancy
ISE 4120 IN-CLASS PROJECT STUDYING REGRESSION
markdown
A Python implementation of John Gruber’s Markdown with Extension support.
pyfolio
Portfolio and risk analytics in Python
research_public
Quantitative research and educational materials
template
A template Sphinx repo
Website
www.harshilpatel.info