annasxu Goto Github PK

crypto-trading-bot

google_trend_data

In this paper, we used time series analysis to analyze data on how frequently “Netflix” is searched for in Google over time. Our findings would allow companies to make better business decisions with the knowledge of when people are likely looking to stay in and watch movies. After preliminary analysis, we found that the data is highly seasonal; hence, we fit a seasonal ARIMA model and determined the best performing model was ARIMA (4,1,3)(1,1,0).

healthyweightregistry

hedging_with_put_options

projects

Project for ORIE 4741

projectsfall2017

Repository for Fall 2017 ORIE 4741 projects.

speed-dating

survival_analysis_relapse_time

The objective of this study is to identify what factors, or combination of factors, contributed to the relapse time of alcoholics. We are also interested in which of the two treatments (cognitive-behavioral or social) is superior in elongating relapse time compared to the other. We reviewed a total of 576 alcoholic patients and used a Cox PH model to analyze a range of genetic and environmental factors. We concluded that family support, prior number of alcohol treatments, type of treatment, age, and the interaction between age and prior number of treatments were significant factors that influenced relapse time. Specifically, stronger family support, more prior treatments, younger ages were associated with a higher hazard of relapse to alcoholism. We also found the treatment emphasizing cognitive behavioral therapy improved relapse time more than social therapy.

topic_modeling_facebook_page_data

weight_loss

This analysis seeks to identify which factors affect middle-aged people’s successful long term weight loss; impulsivity, amount of daily meditation, whether or not the subject kept a food diary, marital status, weight loss goal, and age are considered. Implementing a logistic regression on the dataset, the final model of significant predictors included age, impulsivity, weight loss goal, amount of meditation, and the interactions among age and amount of meditation, impulsivity and amount of meditation, and weight loss goal and impulsivity. Of these factors, higher weight loss goals, low impulsivity, and meditating a medium amount per day all contributed positively to weight loss. For all ages in this “middle-aged” category, meditating a low or medium amount per day also contributed positively to weight loss; however, for subjects with low impulsivity, meditating a low or medium amount per day contributed negatively to weight loss.