arXiv subjects

Zhifeng Ma

Publications and source records attributed to Zhifeng Ma.

5 recordsLinked to original sources

arXiv2025

Gödel Incompleteness Theorem for PAC Learnable Theory from the view of complexity measurement

Different from the view that information is objective reality, this paper adopts the idea that all information needs to be compiled by the interpreter before it can be observed. From the traditional complexity definition, this paper defines the complexity under "the interpreter", which means that heuristically finding the best interpreter is equivalent to using PAC to find the most suitable interpreter. Then we generalize the observation process to the formal system with functors, in which we give concrete proof of the generalized Gödel incompleteness theorem which indicates that there are some objects that are PAC-learnable, but the best interpreter is not found among the alternative interpreters. A strong enough machine algorithm cannot be interpretable in the face of any object. There are always objects that make a strong enough machine learning algorithm uninterpretable, which puts an upper bound on the generalization ability of strong AI.

cs.LO↗

arXiv2024

MS-RNN: A Flexible Multi-Scale Framework for Spatiotemporal Predictive Learning

Spatiotemporal predictive learning, which predicts future frames through historical prior knowledge with the aid of deep learning, is widely used in many fields. Previous work essentially improves the model performance by widening or deepening the network, but it also brings surging memory overhead, which seriously hinders the development and application of this technology. In order to improve the performance without increasing memory consumption, we focus on scale, which is another dimension to improve model performance but with low memory requirement. The effectiveness has been widely demonstrated in many CNN-based tasks such as image classification and semantic segmentation, but it has not been fully explored in recent RNN models. In this paper, learning from the benefit of multi-scale, we propose a general framework named Multi-Scale RNN (MS-RNN) to boost recent RNN models for spatiotemporal predictive learning. We verify the MS-RNN framework by thorough theoretical analyses and exhaustive experiments, where the theory focuses on memory reduction and performance improvement while the experiments employ eight RNN models (ConvLSTM, TrajGRU, PredRNN, PredRNN++, MIM, MotionRNN, PredRNN-V2, and PrecipLSTM) and four datasets (Moving MNIST, TaxiBJ, KTH, and Germany). The results show the efficiency that RNN models incorporating our framework have much lower memory cost but better performance than before. Our code is released at \url{https://github.com/mazhf/MS-RNN}.

cs.CV↗

arXiv2024

MS-LSTM: Exploring Spatiotemporal Multiscale Representations in Video Prediction Domain

The drastic variation of motion in spatial and temporal dimensions makes the video prediction task extremely challenging. Existing RNN models obtain higher performance by deepening or widening the model. They obtain the multi-scale features of the video only by stacking layers, which is inefficient and brings unbearable training costs (such as memory, FLOPs, and training time). Different from them, this paper proposes a spatiotemporal multi-scale model called MS-LSTM wholly from a multi-scale perspective. On the basis of stacked layers, MS-LSTM incorporates two additional efficient multi-scale designs to fully capture spatiotemporal context information. Concretely, we employ LSTMs with mirrored pyramid structures to construct spatial multi-scale representations and LSTMs with different convolution kernels to construct temporal multi-scale representations. We theoretically analyze the training cost and performance of MS-LSTM and its components. Detailed comparison experiments with twelve baseline models on four video datasets show that MS-LSTM has better performance but lower training costs.

cs.CV↗

arXiv2010

Self doping effect and successive magnetic transitions in superconducting Sr$_2$VFeAsO$_3$

We have studied a quinary Fe-based superconductor Sr$_2$VFeAsO$_3$ by the measurements of x-ray diffraction, x-ray absorption, Mössbauer spectrum, resistivity, magnetization and specific heat. This apparently undoped oxyarsenide is shown to be self doped via electron transfer from the V$^{3+}$ ions. We observed successive magnetic transitions within the VO$_2$ layers: an antiferromagnetic transition at 150 K followed by a weak ferromagnetic transition at 55 K. The spin orderings within the VO$_2$ planes are discussed based on mixed valence of V$^{3+}$ and V$^{4+}$.

cond-mat.supr-con↗

arXiv2010

Structural and superconducting properties in LaFeAs1-xSbxO1-yFy

We report the antimony (Sb) doping effect in a prototype system of iron-based supercon-ductors LaFeAsO1-yFy (y=0, 0.1, 0.15). X-ray powder diffraction indicates that the lattice pa-rameters increase with Sb content within the doping limit. Rietveld structural refinements show that, with the partial substitution of Sb for As, while the thickness of the Fe2As2 layers increases significantly, that of the La2O2 layers shrinks simultaneously. So a negative chemical pressure is indeed "applied" to the superconducting-active Fe2As2 layers, in con-trast to the effect of positive chemical pressure by the phosphorus doping. Electrical resis-tance and magnetic susceptibility measurements indicate that, while the Sb doping hardly influences the SDW anomaly in LaFeAsO, it recovers SDW order for the optimally-doped sample of y=0.1. In the meantime, the superconducting transition temperature can be raised up to 30 K in LaFeAs1-xSbxO1-yFy with x=0.1 and y=0.15. The Sb doping effects are discussed in term of both J1-J2 model and Fermi Surface (FS) nesting scenario.

cond-mat.supr-con↗