Tanat Tonguthaisri, CISSP®’s Post

Extracting high-order cosmological information in galaxy surveys with power spectra -Abstract :"The reconstruction method was proposed more than a decade ago to boost the signal of baryonic acoustic oscillations measured in galaxy redshift surveys, which is one of key probes for dark energy. After moving the observed overdensities in galaxy surveys back to their initial position, the reconstructed density field is closer to a linear Gaussian field, with higher-order information moved back into the power spectrum. We find that by jointly analysing power spectra measured from the pre- and post-reconstructed galaxy samples, higher-order information beyond the 2-point power spectrum can be efficiently extracted, which generally yields an information gain upon the analysis using the pre- or post-reconstructed galaxy sample alone. This opens a window to easily use higher-order information when constraining cosmological models. " https://lnkd.in/eCRj_i7w

This paper proposes a new #Riemannian #Generalized #Gaussian #distribution (#RGGD) on that space. The major contributions of this paper are, first of all, providing the exact expression of the #probability #density #function (#PDF) of the RGGD model, as well as an exact expression of the normalizing factor. Furthermore, an estimation of parameters is given using the maximum likelihood of this distribution. The second contribution involves #exploiting the #second-#order statistics of feature maps derived from the first layers of #deep #convolutional #neural #networks (#DCNNs) through the RGGD stochastic model in an image classification framework.----Zakariae Abbad, Ahmed Drissi el maliani, Mohammed El Hassouni, Mohamed Tahar KADAOUI ABBASSI, @Lionel Bombrun, yannick berthoumieu More details can be found at this link: https://lnkd.in/gN8t9SJJ

ICLR2024 Spotlight🌟🌟🌟 PULSE: Physics-based Universal humanoid motion Latent SpacE/Representation Code: github.com/ZhengyiLuo/PUL… Site: zhengyiluo.com/PULSE Paper: arxiv.org/abs/2310.04582 As a motion representation, PULSE is low dimensional (32), high coverage (99.8% of AMASS motion), can speed up downstream task w/ hierarchical RL, and can be randomly sampled as a generative model. PULSE is akin to a foundation model for simulated humanoid control where downstream tasks ranging from simple locomotion, complex terrain traversal, to free-form motion tracking can all reuse this representation.

💥 #ENSI - ENSEMBLE SPACE INVERSION... 💥 👉 This is going to be huge!!! #ENSI will be explained by John Doherty and applied in #Module3, current School edition! Ensemble space inversion (#ENSI) comes with #PEST_HP v.18. It enables efficient, #regularisation-#constrained #calibration of complex, highly-parameterised models. Estimates of parameter and predictive #uncertainty are provided at an unprecedent minimal numerical cost. Tutorials https://lnkd.in/dtwydwkm Video https://lnkd.in/ddKhgYdT Download with #PEST_HP v18 https://lnkd.in/d5v3_vYM

Predicting Dark Photons Using the Dodecahedron Linear String Field Hypothesis: Insights from Collider Data SSRN @SSRN This paper describes a theoretical framework based on the Dodecahedron Linear String Field Hypothesis (DLSFH) to predict the existence of dark photons in (γ) high-energy proton-proton collisions. Specifically, we explore the decay of Higgs bosons produced through the production mode, leveraging the total Run-2 integrated luminosity of 139 recorded by the ATLAS detector at the Large Hadron Collider (LHC). The transverse mass distribution of the system, comprising the photon and the missing transverse momentum from the non-interacting, peaks near the Higgs γ boson mass, providing a distinctive signature. Using data-driven techniques to estimate background processes and a Boosted Decision Tree (BDT) to enhance sensitivity. https://lnkd.in/dS9S-pH2

How do time and space perception compare in terms of reliance on context? My first first-author paper is now online. Special thanks to Mehdi Sanayei . Same principle, but different computations in representing time and space https://lnkd.in/dwgfzthE

Here are some additional physicist-level content ideas with a more technical and calculative focus: 1. _Calculating the beta function of a quantum field theory using the renormalization group equation_: "β(g) = ... #QFT #RenormalizationGroup" 2. _Deriving the Navier-Stokes-Fourier equations from the Boltzmann transport equation_: "∇·v = 0, ∂v/∂t + ... #FluidDynamics #BoltzmannTransport" 3. _Computing the entanglement entropy of a quantum system using the replica trick_: "S_A = ... #QuantumInformation #EntanglementEntropy" 4. _Solving the Einstein-Maxwell equations for a charged black hole_: "Rμν - 1/2Rgμν = ... #GeneralRelativity #EinsteinMaxwell" 5. _Calculating the scattering amplitude of a quantum field theory using the Feynman rules_: "M = ... #QFT #ScatteringAmplitude" 6. _Deriving the Kadanoff-Baym equations from the Schwinger-Dyson equation_: "∂G/∂t + ... #QuantumFieldTheory #KadanoffBaym" 7. _Computing the topological invariant of a topological insulator using the Berry curvature_: "Chern number = ... #TopologicalInsulators #BerryCurvature" 8. _Solving the Dirac-Born-Infeld equation for a nonlinear electromagnetic field_: "∂F/∂t + ... #NonlinearElectrodynamics #DiracBornInfeld" 9. _Calculating the partition function of a quantum system using the transfer matrix method_: "Z = ... #QuantumStatistics #TransferMatrix" 10. _Deriving the Einstein-Hilbert action from the Palatini formalism_: "S = ... #GeneralRelativity #EinsteinHilbert" Let me know if you'd like more ideas or help with anything else!

Interesting! After a one-pixel adversarial attack, presenting a super-weight model collapse, here’s an intriguing insight: The colossal networks {Llama, Llama2, OlMoE, Phi-3} will start outputting gibberish if you set JUST a single weight (super weight) in a 13B Llama2 model to 0. Remarkable OBSERVATION from a model compression perspective! Christopher Angelini, PhD, it aligns with your recent observations, where only <5% of the parameters were important. Paper: https://lnkd.in/gsSq5jnw Enjoy!...

Next week, I will present the latest study based on our work on output-based whirl flutter prediction at the 6th Decennial Aeromechanics Specialists' Conference at the Transformative Vertical Flight Meeting in Santa Clara, CA. Excited to attend this event for the first time as my group continues growing in the area of vertical lift. Preprint here: https://lnkd.in/dtGVFD6Z This paper explores the accuracy of various output-based whirl flutter prediction methods for a propeller-nacelle system with smooth structural nonlinear effects. While conventional methods lose accuracy when strong nonlinearities are present, the method we are investigating maintains consistent accuracy across levels of nonlinearity, excitation amplitudes, and pre-flutter forward speeds.

Of all the research I've done so far, I'm most excited about the project that we've just uploaded to arXiv, about extending persistent homology (degree 0 for now) to the periodic setting: https://lnkd.in/dt9t9bVD I hope our new method will be used to discover exciting insights into material science data, which is often periodic, and hence was challenging to handle by Topological Data Analysis methods in the past. Our topological descriptor is fast to compute (log-linear running time), invariant under choices of how to represent a crystalline material, and stable under perturbations.