Earn your Master's in Data Science online from Syracuse in as few as 18 months. Bachelor's required. https://t.co/Fg8av4vTiV
〽️Weekly Seminars of Condensed Matter and Statistical Physics Group
📝Statistical physics of medical diagnostics
فیزیک آماری مسألهی تشخیص بیماری
🗣 Speaker:
Dr. Abolfazl Ramezanpour
Department of Physics, Shiraz University
⏱Wednesday, September 12, 2018 (۲۱ شهریور)
14:00-15:00
🏛Seminar Room (classroom A), Farmanieh Building, IPM
✍🏻Abstract:
Statistical physics has been widely used to extract macroscopic properties of a wide range of systems from their microscopic interaction models, yet it has not been employed to medical diagnostics. Given an initial subset of observed signs (symptoms, clinical and laboratory findings) with some prior knowledge about the patient (or a complex system like a biological cell), a diagnosis problem simply asks for the most probable diseases (or macrostates with specific phenotypes). An efficient and accurate diagnostic procedure is important specially in the early stages of diseases, where the number and quality of medical evidences are often insufficient to reach a definite diagnosis. Here, we use approximate inference and optimization algorithms of statistical physics to show that a simulation (extrapolation) of the diagnostic process (without doing any real observation) could be helpful as a heuristic strategy in the study of diagnosis problems.
🌐 @ComplexSys
📝Statistical physics of medical diagnostics
فیزیک آماری مسألهی تشخیص بیماری
🗣 Speaker:
Dr. Abolfazl Ramezanpour
Department of Physics, Shiraz University
⏱Wednesday, September 12, 2018 (۲۱ شهریور)
14:00-15:00
🏛Seminar Room (classroom A), Farmanieh Building, IPM
✍🏻Abstract:
Statistical physics has been widely used to extract macroscopic properties of a wide range of systems from their microscopic interaction models, yet it has not been employed to medical diagnostics. Given an initial subset of observed signs (symptoms, clinical and laboratory findings) with some prior knowledge about the patient (or a complex system like a biological cell), a diagnosis problem simply asks for the most probable diseases (or macrostates with specific phenotypes). An efficient and accurate diagnostic procedure is important specially in the early stages of diseases, where the number and quality of medical evidences are often insufficient to reach a definite diagnosis. Here, we use approximate inference and optimization algorithms of statistical physics to show that a simulation (extrapolation) of the diagnostic process (without doing any real observation) could be helpful as a heuristic strategy in the study of diagnosis problems.
🌐 @ComplexSys
Complex Systems Studies
Photo
Interested in Information processing in biological or social systems? Quantum information? Brain-inspired computing? Open Call for students with interest in obtaining a Ph.D. on these topics. Submissions until 21/9. More detailed information available in: https://t.co/LFxe2EgggI
ifisc.uib-csic.es
Open positions for Pre-Docs in 2018/2019
Within the Unit of Excellence María de Maeztu award, IFISC STRATEGIC OBJECTIVES for the period 2018-2022 are the study, exploration, ...
📝 Attack Tolerance of Link Prediction Algorithms: How to Hide Your Relations in a Social Network
Study of the ability to alter connections so as to increase the probability that some connections remain unidentified by link prediction algorithms.
🌐 https://arxiv.org/abs/1809.00152
🎲 @ComplexSys
Study of the ability to alter connections so as to increase the probability that some connections remain unidentified by link prediction algorithms.
🌐 https://arxiv.org/abs/1809.00152
🎲 @ComplexSys
📝 "Mathematical models of collective dynamics and self-organization" (by P. Degond):
"In this paper, we begin by reviewing a certain number of mathematical challenges posed by the modelling of collective dynamics and self-organization. Then..."
🌐 https://arxiv.org/abs/1809.02808
🎲 @ComplexSys
"In this paper, we begin by reviewing a certain number of mathematical challenges posed by the modelling of collective dynamics and self-organization. Then..."
🌐 https://arxiv.org/abs/1809.02808
🎲 @ComplexSys
This is how Google Translate is trained to be biased in gender neutral languages like Persian.
And this paper shows that the bias is even more than the real data!
https://t.co/2TVXLNTgqt
And this paper shows that the bias is even more than the real data!
https://t.co/2TVXLNTgqt
🎊 Great summary of Python basics for scientific computing.
http://cs231n.github.io/python-numpy-tutorial/
http://cs231n.github.io/python-numpy-tutorial/
cs231n.github.io
Python Numpy Tutorial (with Jupyter and Colab)
Course materials and notes for Stanford class CS231n: Deep Learning for Computer Vision.
🌄 Are mountains fractal? Actually there are two scaling regimes (same on earth and venus!)
http://necsi.edu/research/multiscale/earth-venus
Smooth at fine scale, rough at large scales with an exponent satisfying the KPZ (Kardar Parisi Zhang) prediction 0.4.
http://necsi.edu/research/multiscale/earth-venus
Smooth at fine scale, rough at large scales with an exponent satisfying the KPZ (Kardar Parisi Zhang) prediction 0.4.
📝Turn off your e-mail and social media to get more done
Distractions are a fundamental aspect of the modern world, but we don’t have to become hermits to avoid them.
🌐 https://www.nature.com/articles/d41586-018-06213-7
🎲 @ComplexSys
Distractions are a fundamental aspect of the modern world, but we don’t have to become hermits to avoid them.
🌐 https://www.nature.com/articles/d41586-018-06213-7
🎲 @ComplexSys
Nature
Turn off your e-mail and social media to get more done
Nature - Distractions are a fundamental aspect of the modern world, but we don’t have to become hermits to avoid them.