Local feature selection in machine learning provides instance-specific explanations by focusing on the most relevant features for each prediction, enhancing the interpretability of complex models. However, such methods tend to produce misleading explanations by encoding additional information in their selections. In this work, we attribute the problem of misleading selections by formalizing the concepts of label and feature leakage. We rigorously derive the necessary and sufficient conditions under which we can guarantee no leakage, and show existing methods do not meet these conditions. Furthermore, we propose the first local feature selection method that is proven to have no leakage called SUWR. Our experimental results indicate that SUWR is less prone to overfitting and combines state-of-the-art predictive performance with high feature-selection sparsity. Our generic and easily extendable formal approach provides a strong theoretical basis for future work on interpretability with reliable explanations.

In CRYPTO’2019, Gohr firstly combined deep learning with differential cryptanalysis and obtained the deeplearning-based differential distinguishers on round-reduced Speck32/64. In this paper, we introduce the repeated experiment techniques to improve the deep-learning-based differential distinguishers in Gohr’s work. To prove the effectiveness of our techniques, we apply our improved deep-learning-based differential distinguisher scheme into round-reduced Simeck32/64 and obtain long to 15-round deep-learning-based differential distinguisher. To the best of our knowledge, this is the longest-round distinguisher for Simeck32/64 when compared with the 14round differential distinguisher obtained by traditional wisdom.

At CRYPTO 2019, A. Gohr made a breakthrough in combining classical cryptanalysis and deep learning and applied his method to round reduced SPECK successfully. However, his suggested neural-based distinguisher scheme is only limited to differential cryptanalysis. In this paper, we have the following contributions: 1. We combine integral cryptanalysis and deep learning to propose our neural-based integral distinguisher scheme for the first time. To illustrate the effectiveness of our distinguisher scheme, we apply it to block ciphers of different structures, such as substitution-permutation structure ciphers (PRESENT and RECTANGLE), Feistel structure cipher (LBLOCK), and add-rotate-XOR cipher (SPECK) and compare the results with the state-of-the-art classical integral distinguishing method, namely, the bit-based division property. To our great surprise, our neural network-based integral distinguisher can extend the number of distinguished rounds for all block ciphers by two additional rounds (except RECTANGLE, where it is improved by one round) under the same data complexity. 2. As an additional advantage of our scheme, we demonstrate that our Neural Distinguisher (ND) is not only helpful for block cipher designers but also can assist attackers to mount key recovery attacks. To this end, we show how to exploit our ND to mount a key recovery attack and apply it to SPECK32/64. Out of the 1000 trials of key recovery attacks with different keys in 45% of cases, the first suggested subkey is exactly the real subkey of the last round of the cipher. For the remaining 55%, the second or third suggested subkey is exactly the real subkey of the last round of the cipher. 3. To have a piece of concrete evidence for the advantage of our scheme over classical integral methods, we design an experiment known as the same-difference experiment. In this experiment, we show that our ND can learn some features beyond the capabilities of classical integral methods. We then propose a set of features that can justify the gap between classical integral methods and our neural-based integral distinguisher and verify them by further experiments.

Abstract Optical character recognition (OCR) is crucial for a deeper access to historical collections. OCR needs to account for orthographic variations, typefaces, or language evolution (i.e., new letters, word spellings), as the main source of character, word, or word segmentation transcription errors. For digital corpora of historical prints, the errors are further exacerbated due to low scan quality and lack of language standardization. For the task of OCR post-hoc correction, we propose a neural approach based on a combination of recurrent (RNN) and deep convolutional network (ConvNet) to correct OCR transcription errors. At character level we flexibly capture errors, and decode the corrected output based on a novel attention mechanism. Accounting for the input and output similarity, we propose a new loss function that rewards the model’s correcting behavior. Evaluation on a historical book corpus in German language shows that our models are robust in capturing diverse OCR transcription errors and reduce the word error rate of 32.3% by more than 89%.

Insulation faults account for a high proportion in the faults of electrical equipment. Insulation defects or faults of electrical equipment may cause excessive temperature rise or partial discharge, which can be used as the criteria of insulation state of electrical equipment. However, the existing detection methods can't meet the requirements of safe and stable operation of modern substations. The exploration of new methods for temperature rise and partial discharge detection has become important for online detection of electrical equipment. Photoelectric sensor is a device that converts optical signal into electrical signal. It can be used to detect the optical signal generated by the running electrical equipment. The infrared photoelectric sensor can detect the temperature of electrical equipment, and the ultraviolet photoelectric sensor can detect the ultraviolet pulse signal generated by partial discharge of electrical equipment. In this paper, the characteristics of infrared photoelectric sensor's temperature changing with the detection distance are studied, and then the optimum detection distance is obtained. The relationship between the output pulse signal of ultraviolet photoelectric sensor and discharge intensity is analyzed, and the attenuation characteristics of pulse signal with the increase of propagation distance are also analyzed. The optimum placement position is selected. An insulation defect detection system for electrical equipment based on infrared and ultraviolet photoelectric sensing technology is constructed. Based on the adaptive fuzzy neural network, the insulation state of electrical equipment is synthetically judged by the signals of the infrared and ultraviolet photoelectric sensors. Experimental results show that through the combined detection of infrared and ultraviolet photoelectric sensors, and then information fusion diagnosis, it can effectively reduce the single sensor's misjudgment caused by one-sided information, and the accuracy of fault diagnosis is significantly improved.

Effective fault diagnosis and parameter estimation is the precondition to ensure safe and reliable operation of Gas Insulated Switchgear (GIS). In this paper, aiming at GIS busbar electrical contact overheating fault, an Improved Kalman Filter (IKF) method is proposed, which can predict the temperature rise of GIS internal contacts accurately. In this method, the state equation and the measurement equation are obtained by the Auto-Regressive Integrated Moving Average (ARIMA) model in time series theory and the Back Propagation Neural Network (BPNN) model respectively, which reduces the accumulated error caused by one single prediction method. Genetic Algorithm (GA) is used to optimize the system noise and measurement noise, which reduces the interference of noise on the prediction result. The experiment and comparison results show that the proposed IKF prediction method can effectively estimate and predict the temperature rise of the contacts in the GIS busbar with high prediction accuracy and strong anti-interference ability.

The Internet of Things has become a spotlight for a long period of time and generates massive amounts of sensor data. Thus, data centers play more and more crucial roles in processing and analyzing the explosively increasing data. To remedy the shortcomings of traditional tree-based structure, many novel server-centric network structures have been proposed in recent years. Their original routing mechanisms based on divide and conquer (DC) are not able to work out the shortest paths. So, there is still promotion room for communication delay reduction. Since dynamic programming (DP) is a classical strategy to obtain optimal solution, this paper proposes a routing mechanism based on DP and applies it to data center for better solving the weakness occurred by DC. Experiments firmly support the conclusion that adopting DP in routing calculation achieves appealing performance of short latency, great fault-tolerance and reasonable resource consumption. Theoretical analysis also proves that it is applicable to most popular structures.