578 research outputs found
Precision Health and AI: improving health for everyone - Arjun Panesar (DDM Health)
No full textThis video is the twelfth talk from our two day Future Blood Testing: Challenges & Opportunities Event that took place on the 14/09/2022.
Precision Health and AI: improving health for everyone - Arjun Panesar (DDM Health)
Bio: Arjun Panesar is the founder of DDM Health, providers of clinically-validated digital health solutions to over 1.8 million people. Benefiting from almost two decades of experience in big data, AI and AI ethics, Arjun leads the development of evidence-based digital innovations that harness the power of machine learning to provide precision medicine to patients, health services, and governments. Arjun’s work has received international recognition featuring in the Forbes, New Scientist, BBC and The Times. Arjun is a best-selling author on the topics of healthcare and AI, authoring two editions of Machine Learning and AI in Healthcare, and contributing to Handbook of Global Health, a major reference work. Arjun is an advisor to the Information School, University of Sheffield, Fellow to the NHS Innovation Accelerator, visiting lecturer at University of Warwick Medical School, and was recognised by Imperial College as an Alumni Leader for his contribution and impact to society.
Further details on this event can be found at: https://futurebloodtesting.org/event/13-14-09-2022/
This video is an output from the Future Blood Testing Network which is funded by EPSRC under Grant Number EP/W000652/1
YouTube Link: https://youtu.be/clPmdeLP5_
Financial Management of Globalization of Developing Countries
Full text linkhuman development, economic growth, globalization, inequality, poverty
Opportunistic overlapping: Joint scheduling of uplink URLLC/eMBB traffic in NOMA based wireless systems
No full textStereo vision-based vehicular proximity estimation
No full textThis thesis describes an innovative and cost effective method to develop a low-cost following distance logging algorithm for volunteer participants which will allow quantitative research in driving behavior. Sparse stereo depth estimation methods along with a license plate localization algorithm has been used in order to achieve this. The depth is estimated by processing the video feeds from the stereo camera setup mounted inside a car looking out of the front window. License plate localization is used as a means to localize the position of the car within the image. Depth is estimated from the disparity, which is calculated using the rectified images of the frames from both the video streams.M.S.Includes bibliographical referencesby Arjun Krishn
Freedom of choice or force of circumstance? : Eastern European sex-workers in the Republic of Cyprus ; paper for the conference 'Alltag der Globalisierung. Perspektiven einer transnationalen Anthropologie', January 16-18, 2003, Institute of Cultural Anthropology and European Ethnology, Johann Wolfgang Goethe University, Frankfurt am Main
No full textThis paper focuses on Eastern European migrants who, since the beginning of the 1990s, are entering the Republic Cyprus as “artistes”. This is a visa permit status as well as an euphemism for short-term work permits in the local sex industry. In addition to exploring the migrational experiences of these women and their living and working conditions in the Republic of Cyprus, the paper reconstructs, empirically and analyt ically, the connection between immigration and the local sex industry. Here, several categories of social actors and institutions in Cyprus are actively involved. The rhetoric of government representatives, entrepreneurs and clients in the sex business on the one hand is contrasted with the discourse of local NGO representatives concerned with immigrants’ rights on the other hand. The paper comes to the conclusion that all of these discursive positions ultimately do not do justice to the complex process of decisionmaking that women undergo who migrate into the sex industry. Either, freedom of choice is emphasized – such as by entrepreneurs and the government – or the domination of women – as in the public statements of the NGO. In order to analyze the ambivalent tension between freedom of choice and submission to force by which the women’s decision is characterized, the author employs Michel Foucault’s concept of governmentality, which describes forms of political regulation that use the individual’s freedom of action as an instrument to exercise power
Fine-tuning generative models
No full textThis electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (pages 73-76).Deep generative models have emerged as a powerful modeling paradigm for making sense of large amounts of unlabeled real-world data. In particular, the representations produced by these models have proven to be useful both in improving human understanding of the factors of variation in the original dataset and in downstream tasks such as classification. Most current algorithms, however, require training a bespoke model from scratch, which can be both expensive and time-consuming. Instead, we propose various methods of fine-tuning pre-trained generative models to achieve these goals, and evaluate these methods quantitatively on few-shot classification and interpretability tasks.by Arjun Khandelwal.M. Eng.M.Eng. Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Scienc
Improving parking garage efficiency using reservation optimization techniques
No full textParking in urban and suburban districts is becoming increasingly problematic due the information gap between the parking garages and the average commuter. This thesis describes and evaluates techniques that can be implemented by parking garages to augment parking garage efficiency. The issues studied in this thesis were i) Real-time tracking of car position ii) maximizing the number of reservations made for the parking garage by re-arrangement of existing reservations (Reservation Defragmentation) and iii) maximizing revenue for the parking garage through increased occupancy (Revenue Management). For the tracking problem, in order to be able to track the real-time position of the vehicle inside the parking garage, we have proposed two techniques. The first one involves a high accuracy algorithm that takes as input a higher number of sensor values (high accuracy) and the second one is a lower cost algorithm that takes as input fewer number of sensor values (low cost). We simulated various conditions of sensor failure rate and determined our metric to be number of tracked points as a percentage of the path to the destination. We determined limitations of these algorithms with respect to maximum speed of cars and inter-car distance. For the reservation defragmentation problem, we looked at increasing occupancy efficiency for i) Next day reservations and ii) Current day reservations. For this problem, we implemented three algorithms. For next day reservations, we established metrics to determine the efficiency of the algorithm including number of free parking spots created and reduction in lengths of free space in between the reservations. For current day reservations, our metric was the increase in maximum occupancy observed due to defragmentation. For increased revenue management, we suggested the application of two techniques: Booking limits and Overbooking. In booking limits, two-fare class of parking was suggested and the number of spots that need to be reserved for higher class (Capacity of garage - booking limit) was determined for probability distributions of customer arrival such as Poisson distribution. Since the practice of overbooking is done in order to compensate for the no-shows that occur despite reservations made, we have suggested an algorithm to determine the amount of overbooking based on a Gaussian distribution of customer ‘no-shows’. We obtained the following results for the algorithms implemented. In case of the tracking algorithm, as the sensor failure rate increased, the inaccuracy of the two proposed algorithms also increased. For 2% failure rate, we track 0.4% of the incoming cars inaccurately (given that a tracking is marked as correct if 75% or less of all sensors along the path of the car fail). In case of reservation defragmentation, we obtained best results for Recursive First-Fit algorithm. For next day reservation defragmentation, using a mean of 15% cancellation of reservations resulted in 14.6% decrease in occupied parking spots which can then lead to increased occupancy and 46.3% decrease in inter-reservation free space sizes for a 1000 arrival reservation system. The reservations were exponentially distributed with a mean of 20 reservations/hour. For current day reservations, we were able to increase maximum occupancy of the parking garage by 5.5% using Recursive First Fit algorithm. Among other conditions, we have evaluated Poisson arrival distribution with corporate arrival rate 100 cars/hour (Flintsch et al., 2006) [56] and corporate fare twice of leisure fare. For this condition, protection level (number of parking slots reserved for corporate class) is determined to be 20% of garage capacity. We also evaluated Binomial distribution with probability of incoming customer to be corporate customer as 0.5 and corporate fare twice of leisure fare. For this condition, protection level is determined to be 50% of garage capacity. We evaluated overbooking for several combinations of No-show rates, mean and standard deviation values and the highest amount of overbooking we obtained was 1.93 times maximum garage capacity and this implies that permitting this number of reservations for the parking garage would minimize the number of parking spots being under-utilized and increase the revenue of the parking garage operator due to effective use of parking spots. The algorithms have been simulated for different arrival distributions (for Revenue Management), different arrival rates (tracking) as well as variable durations of stay (reservation defragmentation). Besides the problems mentioned, there are certain other aspects, such as generalizing the tracking algorithms for parking garages of arbitrary layouts represents the work that needs to be done in the future.M.S.Includes bibliographical referencesby Arjun Ra
Raw data for "Activation of oligonucleotide polyanions using collisions, electrons and photons in a timsOmni platform"
No full textRaw mass spectral data used to prepare the figures of the article "Activation of oligonucleotide polyanions using collisions, electrons and photons in a timsOmni platform", by Frédéric Rosu,1 Rim Chiba,1 Arjun Mani Mallika,1 Athanasios Smyrnakis,2 Jean-François Greisch,3 Dimitris Papanastasiou,2 Valérie Gabelica1*
1 School of Pharmaceutical Sciences, University of Geneva, 1205 Geneva, Switzerland.
2 Fasmatech Science & Technology, 15232 Chalandri, Athens, Greece.
3 Bruker Switzerland AG, 8117 Fällanden, Switzerland.
* Corresponding author: [email protected]
The data is organized by Table or Figure number as in the manuscript main text and supporting information
Recommended from our members
Schedulers for next generation wireless networks : realizing QoE trade-offs for heterogeneous traffic mixes
Full text linkIn this thesis we will focus on the design of schedulers for next generation wireless networks which support application mixes, characterized by different, possibly complex, application/user Quality of Experience (QoE) metrics. The central problem underlying resource allocation for such systems is realizing QoE trade-offs among various applications/users given the dynamic loads and capacity variability they would typically see. In the first part of the thesis our focus is on applications where QoE depends on flow-level delay-based metrics. We consider system-wide metrics which directly capture both users' QoE metrics and appropriate QoE trade-offs among various applications for a wide range of system loads. This approach is different from the traditional wireless scheduler designs which have been driven by rate-based criteria, e.g., utility maximizing/proportionally fair, and/or queue-based packet schedulers which do not directly reflect the link between flow-level delays and users' QoE. In the second part of this thesis we address the key design challenges in networks supporting Ultra Reliable Low Latency Communications (URLLC) traffic which requires extremely high reliability (99.999%) and very low delays (1 msec). We will explore three different types flow delay-based metrics in this proposal, based on 1) overall mean delay; 2) functions of mean delays; and, 3) mean of functions of delays. We begin by considering minimization of mean flow delay for an M/GI/1 queuing model for a wireless Base Station (BS) where the flow size distributions are of the New Better than Used in Expectation + Decreasing Hazard Rate (NBUE +DHZ) type. Such a flow size distribution have been observed in real systems and we too validate this model based on collected data. Using a combination of analysis and simulation we show that our scheduler achieves good performance for users that might correspond to interactive applications like web browsing and/or stored video streaming and is robust to variations in system loads. Next we consider a generalization of this approach where we minimize a metric based on cost functions of the mean flow delays in a multi-class system where users/flows are classified based on their respective QoE requirements and each class's QoE requirement is modeled by its respective cost function. This approach helps us model QoE more accurately and gives us more flexibility in considering QoE trade-offs among heterogeneous user classes. We optimize two different metrics based on how we average the cost functions of delays, namely, functions of mean delays; and mean of functions of delays. The former can be used when users' experiences are sensitive to mean delays and while the latter can be used when user's experience is also sensitive to higher moments of delays, e.g., variance or soft thresholds on delay. Extensive simulations confirm the effectiveness of our proposed approaches at realizing various QoE trade-offs and performance. In 5G wireless networks URLLC traffic is expected to support many applications like industrial automation, mission critical traffic, virtual traffic etc, where the wireless network has to reliability transport small packets with very high reliability and low delays. We address the following aspects related to the system design for URLLC traffic, 1) quantifying the impact of various system parameters like system bandwidth, link SINR, delay and latency constraints on URLLC 'capacity'; 2) provisioning wireless system appropriately to meet URLLC Quality of Service (QoS) requirements; and, 3) designing efficient Hybrid Automatic Repeat Request (HARQ) schemes for transmitting small packets. Further, due the heterogeneity in delay requirements between URLLC and other types of traffic, sharing radio resources between them creates its own unique challenges. We develop efficient multiplexing schemes between URLLC traffic and other mobile broadband traffic based on preemptive puncturing/superposition of the mobile broadband transmissions by URLLC transmissions.Electrical and Computer Engineerin
C4 the original pre-workout & explosive force
Full text linkC4 pre-workout can potentially increase force generated through its properties in the beverage, this has resulted in billions of servings of C4 pre-workout sold worldwide. Caffeine that is ingested 30 to 90 minutes prior to exercise has been shown to result in performance increases of up to 6% in events lasting from a few minutes to several hours. (Glaister & Gissane, 2018) Caffeine enhances peak power production, improves cognitive performance and enhances readiness to invest physical effort. (Duncan et al., 2019) Purpose: To determine if there is a significant difference in force when the C4 supplement is used vs. when our placebo is used.Not peer reviewedStudent Research Day Poster (2019
- …
