-
FastMapSVM: Classifying Complex Objects Using the FastMap Algorithm and Support-Vector Machines
Authors:
Malcolm C. A. White,
Kushal Sharma,
Ang Li,
T. K. Satish Kumar,
Nori Nakata
Abstract:
Neural Networks and related Deep Learning methods are currently at the leading edge of technologies used for classifying objects. However, they generally demand large amounts of time and data for model training; and their learned models can sometimes be difficult to interpret. In this paper, we advance FastMapSVM -- an interpretable Machine Learning framework for classifying complex objects -- as…
▽ More
Neural Networks and related Deep Learning methods are currently at the leading edge of technologies used for classifying objects. However, they generally demand large amounts of time and data for model training; and their learned models can sometimes be difficult to interpret. In this paper, we advance FastMapSVM -- an interpretable Machine Learning framework for classifying complex objects -- as an advantageous alternative to Neural Networks for general classification tasks. FastMapSVM extends the applicability of Support-Vector Machines (SVMs) to domains with complex objects by combining the complementary strengths of FastMap and SVMs. FastMap is an efficient linear-time algorithm that maps complex objects to points in a Euclidean space while preserving pairwise domain-specific distances between them. We demonstrate the efficiency and effectiveness of FastMapSVM in the context of classifying seismograms. We show that its performance, in terms of precision, recall, and accuracy, is comparable to that of other state-of-the-art methods. However, compared to other methods, FastMapSVM uses significantly smaller amounts of time and data for model training. It also provides a perspicuous visualization of the objects and the classification boundaries between them. We expect FastMapSVM to be viable for classification tasks in many other real-world domains.
△ Less
Submitted 15 June, 2022; v1 submitted 7 April, 2022;
originally announced April 2022.
-
Domain-Specific Compilers for Dynamic Simulations of Quantum Materials on Quantum Computers
Authors:
Lindsay Bassman,
Sahil Gulania,
Connor Powers,
Rongpeng Li,
Thomas Linker,
Kuang Liu,
T. K. Satish Kumar,
Rajiv K. Kalia,
Aiichiro Nakano,
Priya Vashishta
Abstract:
Simulation of the dynamics of quantum materials is emerging as a promising scientific application for noisy intermediate-scale quantum (NISQ) computers. Due to their high gate-error rates and short decoherence times, however, NISQ computers can only produce high-fidelity results for those quantum circuits smaller than some given circuit size. Dynamic simulations, therefore, pose a challenge as cur…
▽ More
Simulation of the dynamics of quantum materials is emerging as a promising scientific application for noisy intermediate-scale quantum (NISQ) computers. Due to their high gate-error rates and short decoherence times, however, NISQ computers can only produce high-fidelity results for those quantum circuits smaller than some given circuit size. Dynamic simulations, therefore, pose a challenge as current algorithms produce circuits that grow in size with each subsequent time-step of the simulation. This underscores the crucial role of quantum circuit compilers to produce executable quantum circuits of minimal size, thereby maximizing the range of physical phenomena that can be studied within the NISQ fidelity budget. Here, we present two domain-specific quantum circuit compilers for the Rigetti and IBM quantum computers, specifically designed to compile circuits simulating dynamics under a special class of time-dependent Hamiltonians. The compilers outperform state-of-the-art general-purpose compilers in terms of circuit size reduction by around 25-30% as well as wall-clock compilation time by around 40% (dependent on system size and simulation time-step). Drawing on heuristic techniques commonly used in artificial intelligence, both compilers scale well with simulation time-step and system size. Code for both compilers is included to enhance the results of dynamic simulations for future researchers. We anticipate that our domain-specific compilers will enable dynamic simulations of quantum materials on near-future NISQ computers that would not otherwise be possible with general-purpose compilers.
△ Less
Submitted 15 April, 2020;
originally announced April 2020.
-
Measuring Territorial Control in Civil Wars Using Hidden Markov Models: A Data Informatics-Based Approach
Authors:
Therese Anders,
Hong Xu,
Cheng Cheng,
T. K. Satish Kumar
Abstract:
Territorial control is a key aspect shaping the dynamics of civil war. Despite its importance, we lack data on territorial control that are fine-grained enough to account for subnational spatio-temporal variation and that cover a large set of conflicts. To resolve this issue, we propose a theoretical model of the relationship between territorial control and tactical choice in civil war and outline…
▽ More
Territorial control is a key aspect shaping the dynamics of civil war. Despite its importance, we lack data on territorial control that are fine-grained enough to account for subnational spatio-temporal variation and that cover a large set of conflicts. To resolve this issue, we propose a theoretical model of the relationship between territorial control and tactical choice in civil war and outline how Hidden Markov Models (HMMs) are suitable to capture theoretical intuitions and estimate levels of territorial control. We discuss challenges of using HMMs in this application and mitigation strategies for future work.
△ Less
Submitted 9 December, 2017; v1 submitted 17 November, 2017;
originally announced November 2017.
-
Poverty Index With Time Varying Consumption and Income Distributions
Authors:
Amit K Chattopadhyay,
T Krishna Kumar,
Sushanta K Mallick
Abstract:
In a recent work (Chattopadhyay, A. K. et al, Europhys. Lett. {\bf 91}, 58003, 2010) based on food consumption statistics, we showed how a stochastic agent based model could represent the time variation of the income distribution statistics in a developing economy, thereby defining an alternative \enquote{poverty index} (PI) that largely agreed with poverty gap index data. This PI used two variabl…
▽ More
In a recent work (Chattopadhyay, A. K. et al, Europhys. Lett. {\bf 91}, 58003, 2010) based on food consumption statistics, we showed how a stochastic agent based model could represent the time variation of the income distribution statistics in a developing economy, thereby defining an alternative \enquote{poverty index} (PI) that largely agreed with poverty gap index data. This PI used two variables, the probability density function of the income statistics and a consumption deprivation (CD) function, representing the shortfall in the minimum consumption needed for survival. Since the time dependence of the CD function was introduced there through data extrapolation only and not through an endogenous time dependent series, this model left unexplained how the minimum consumption needed for survival varies with time. The present article overcomes these limitations and arrives at a new unified theoretical structure through time varying consumption and income distributions where trade is only allowed when the income exceeds consumption deprivation (CD). Our results reveal that such CD-dynamics reduces the threshold level of consumption of basic necessities, suggesting a possible dietary transition in terms of lower saturation level of food-grain consumption. The new poverty index conforms to recently observed trends more closely than conventional measures of poverty and allows probabilistic prediction of PI for future times.
△ Less
Submitted 18 August, 2016;
originally announced August 2016.