Description: With the help of a very robust largescale crawling framework, we aggregate all the real-time and historical (millions of) Swiss and French housing advertisements for rent, available anywhere on the internet. With the state of the art algorithms, we provide customized suggestions to our users to help them find the rooms as well as connect with potential roommates. We use two-sided market algorithms for optimum matching. The working principle of the app is similar to the Tinder app.
Description: We developed a framework for analyzing regional level incidence data for ten countries. We quantify the impact of lockdowns on epidemic progression. With the framework, it is possible to quantify the spatial diffusion of the epidemic process and the factors affecting the epidemic growth.
Description: We use the historical quarterly fundamental ratios (for all S&P 500 companies) to design a portfolio construction algorithm using machine learning methods and copula techniques to show a significant improvement compared to the S&P 500 benchmarks.
Description: We use the high-frequency forex data to demonstrate the presence of exploitable self-excitation and positive feedbacks. We design a machine-learning trading algorithm to exploit the effects and show that the algorithm is in fact profitable with an impressive Sharpe ratio.
Description: Using an optical speckle analysis technique, we show that predictability of the instability of the crack growth is possible before the sample-dependent minimum strain rate because of the growth of the fracture process zone.
Description: We analyzed the complete history of international cricket containing hundreds of thousands of performances over a period of 142 years to validate the presence of hot-hand effects (success triggering future success) in player performances using a self-excited point process model.
Description: An unified evolutionary decision-making framework to model the emergence of wealth inequality under various wealth visibility conditions.
Quantifying social spillovers in scientific careers
Status:
Description: Using a massive dataset containing hundreds of millions of citation records in scientific careers, we quantify the social influence driving the individual scientific success.
Defining and quantifying the maturation of the financial bubbles
Status:
Description: We decluster the financial bubbles using multifractal random walks and by quantifying the accelerated patterns of the price movement. Using the growth parameter, we predict the crash behavior.
Tipping points in technology adoption and social convention in the presence of the law of small numbers
Status:
Description: We justify the evolutionary selection of cognitive biases by showing that society under cognitive biases converge and adapt to the social changes much faster than the people following the Nash strategy.
Phase transition in risk-taking ability and wealth inequality within a risk-averse society in the presence of social influence
Status:
Description: Keeping the social influence as a control parameter, we show a phase transition in wealth inequality and risk-taking abilities in a risk-averse society.
Uncovering the socio-economic contagion during the cryptocurrency bubbles
Status:
Description: With the help of a number of social and financial datasets, we quantify the bivariate positive feedback effects during the cryptocurrency bubbles.
Follow Me!