Timothy DeLise's research while affiliated with Université de Montréal and other places

Publications (8)

Preprint
Full-text available
High frequency financial data is burdened by a level of randomness that is unavoidable and obfuscates the task of modelling. This idea is reflected in the intraday evolution of limit orders book data for many financial assets and suggests several justifications for the use of stochastic models. For instance, the arbitrary distribution of inter arri...
Chapter
This research proposes a data segmentation algorithm which combines t-SNE, DBSCAN, and Random Forest classifier to form an end-to-end pipeline that separates data into natural clusters and produces a characteristic profile of each cluster based on the most important features. Out-of-sample cluster labels can be inferred, and the technique generaliz...
Preprint
This research investigates pricing financial options based on the traditional martingale theory of arbitrage pricing applied to neural SDEs. We treat neural SDEs as universal It\^o process approximators. In this way we can lift all assumptions on the form of the underlying price process, and compute theoretical option prices numerically. We propose...
Preprint
Full-text available
This research proposes a data segmentation technique which is easy to interpret and generalizes well. The technique combines t-SNE, DBSCAN, and Random Forest classifier algorithms to form an end-to-end pipeline that separates data into natural clusters and produces a characteristic profile of each cluster based on the most important features. Out-o...
Article
Background: Technology-based computational strategies that leverage social network site (SNS) data to detect substance use are promising screening tools but rely on the presence of sufficient data to detect risk if it is present. A better understanding of the association between substance use and SNS participation may inform the utility of these t...
Preprint
BACKGROUND Technology-based computational strategies that leverage social network site (SNS) data to detect substance use are promising screening tools but rely on the presence of sufficient data to detect risk if it is present. A better understanding of the association between substance use and SNS participation may inform the utility of these tec...
Article
Full-text available
Social media may provide new insight into our understanding of substance use and addiction. In this study, we developed a deep-learning method to automatically classify individuals’ risk for alcohol, tobacco, and drug use based on the content from their Instagram profiles. In total, 2287 active Instagram users participated in the study. Deep convol...

Citations

... With t-SNE, we condense the 300-dimensional semantic space into a two-dimensional map, which allows for a straightforward visual inspection and labeling of semantic sub-categories. Alternatively, additional clustering algorithms, such as k-means, could be considered atop the t-SNE visualization (Taskesen and Reinders, 2016;Devassy et al., 2020;DeLise, 2021). However, some concerns have been presented on utilizing another clustering algorithm with t-SNE (van der Maaten and Hinton, 2008). ...
... Communication records can convey information from social networks. Bergman et al. [22] found that compared with people not using addictive substances, anyone who is at risk of drinking or taking drugs (e.g., marijuana, cocaine, and heroin) are more likely to post on Instagram. Wearable devices, such as watches, wristbands, chest straps, and transdermal patches, facilitate real-time collection of various types of data, such as biological samples, location, and physiological changes, to help researchers understand the causes and consequences of substance use [19] . ...
... Thanks to almost 3.8 billion users across the globe [4], social media platforms are a precious data source and researchers have often profitably analysed comments extracted from e.g. Facebook [5], Twitter [6,7], Instagram [8], etc. regarding political, business and healthcare issues. ...