Dynamical Mechanism of Sampling-Based Probabilistic Inference under Probabilistic Population Codes

Animals make efficient probabilistic inferences based on uncertain and noisy information from the outside environment. It is known that probabilistic population codes, which have been proposed as a neural basis for encoding probability distributions, allow general neural networks (NNs) to perform near-optimal point estimation. However, the mechanism of sampling-based probabilistic inference has not been clarified. In this study, we trained two types of artificial NNs, feedforward NN (FFNN) and recurrent NN (RNN), to perform sampling-based probabilistic inference. Then we analyzed and compared their mechanisms of sampling. We found that sampling in RNN was performed by a mechanism that efficiently uses the properties of dynamical systems, unlike FFNN. In addition, we found that sampling in RNNs acted as an inductive bias, enabling a more accurate estimation than in maximum a posteriori estimation. These results provide important arguments for discussing the relationship between dynamical systems and information processing in NNs.

The Cyclical Behavior of Unemployment and Wages under Information Frictions

I propose a new mechanism for sluggish wages based on workers’ noisy information about the state of the economy. Wages do not respond immediately to a positive aggregate shock because workers do not (yet) have enough information to demand higher wages. The model is robust to two major criticisms of existing theories of sluggish wages and volatile unemployment, namely, that wages are flexible for new hires and the flow opportunity cost of employment (FOCE) is pro-cyclicality. The model generates volatility in the labor market as well as wage and FOCE elasticities with respect to productivity, consistent with the data. (JEL E24, E32, J24, J31, J63)

A comparative study using supervised learning for anomaly detection in network traffic

A user connects to hundreds of remote networks daily, some of which can be corrupted by malicious sources. To overcome this problem, a variety of Network Intrusion Detection systems are built, which aim to detect harmful networks before they establish a connection with the user’s local system. This paper focuses on proposing a model for Anomaly based Network Intrusion Detection systems (NIDS), by performing comparisons of various Supervised Learning Algorithms on metric of their accuracy. Two datasets were used and analysed, each having different properties in terms of the volume of data they contain and their use cases. Feature engineering was done to retrieve the most optimum features of both the datasets and only the top 25% best features were used to build the models – a smaller subset of features not only aids in decreasing the capital required to collect the data but also gets rid of redundant and noisy information. Two different splicing methods were used to train the data and each method showed different trends on the ML models.

COVIDSensing: Social Sensing Strategy for the Management of the COVID-19 Crisis

The management of the COVID-19 pandemic has been shown to be critical for reducing its dramatic effects. Social sensing can analyse user-contributed data posted daily in social-media services, where participants are seen as Social Sensors. Individually, social sensors may provide noisy information. However, collectively, such opinion holders constitute a large critical mass dispersed everywhere and with an immediate capacity for information transfer. The main goal of this article is to present a novel methodological tool based on social sensing, called COVIDSensing. In particular, this application serves to provide actionable information in real time for the management of the socio-economic and health crisis caused by COVID-19. This tool dynamically identifies socio-economic problems of general interest through the analysis of people’s opinions on social networks. Moreover, it tracks and predicts the evolution of the COVID-19 pandemic based on epidemiological figures together with the social perceptions towards the disease. This article presents the case study of Spain to illustrate the tool.

Tracking Unmanned Aerial Vehicles Based on the Kalman Filter Considering Uncertainty and Error Aware

Recently, Unmanned Aerial Vehicles (UAVs) have made significant impacts on our daily lives with the advancement of technologies and their applications. Tracking UAVs have become more important because they not only provide location-based services, but are also faced with serious security threats and vulnerabilities. UAVs are smaller in nature, move with high speed, and operate in a low-altitude environment, which makes it conceivable to track UAVs using fixed or mobile radars. Kalman Filter (KF)-based methodologies are widely used for extracting valuable trajectory information from samples composed of noisy information. As UAVs’ trajectories resemble uncertain behavior, the traditional KF-based methodologies have poor tracking accuracy. Recently, the Diffusion-Map-based KF (DMK) was introduced for modeling uncertainties in the environment without prior knowledge. However, the model has poor accuracy when operating in environments with higher noise. In order to achieve better tracking performance, this paper presents the Uncertainty and Error-Aware KF (UEAKF) for tracking UAVs. The UEAKF-based tracking method provides a good tradeoff among preceding estimate confidence and forthcoming measurement under dynamic environments; the resulting filter is robust and nonlinear in nature. The experimental results showed that the UEAKF-based UAV tracking model achieves much better Root Mean Square Error (RMSE) performance compared to the existing particle filter-based and DMK-based UAV tracking models.

Localization Algorithm for 3D Sensor Networks: A Recursive Data Fusion Approach

Location-based applications for security and assisted living, such as human location tracking, pet tracking and others, have increased considerably in the last few years, enabled by the fast growth of sensor networks. Sensor location information is essential for several network protocols and applications such as routing and energy harvesting, among others. Therefore, there is a need for developing new alternative localization algorithms suitable for rough, changing environments. In this paper, we formulate the Recursive Localization (RL) algorithm, based on the recursive coordinate data fusion using at least three anchor nodes (ANs), combined with a multiplane location estimation, suitable for 3D ad hoc environments. The novelty of the proposed algorithm is the recursive fusion technique to obtain a reliable location estimation of a node by combining noisy information from several nodes. The feasibility of the RL algorithm under several network environments was examined through analytic formulation and simulation processes. The proposed algorithm improved the location accuracy for all the scenarios analyzed. Comparing with other 3D range-based positioning algorithms, we observe that the proposed RL algorithm presents several advantages, such as a smaller number of required ANs and a better position accuracy for the worst cases analyzed. On the other hand, compared to other 3D range-free positioning algorithms, we can see an improvement by around 15.6% in terms of positioning accuracy.

Molecular convolutional neural networks with DNA regulatory circuits

Complex biomolecular circuits enable cells with intelligent behavior for survival before neural brains evolved. Synthesized DNA circuits in liquid phase developed as computational hardware can perform neural-network-like computation that harness the collective properties of complex biochemical systems, however the scaling up in complexity remains challenging to support more powerful computation. we present a systematic molecular implementation of the convolutional neural network (ConvNet) algorithm with synthetic DNA regulatory circuits based on a simple DNA switching gate architecture. We experimentally demonstrated that a DNA-based ConvNet based on shared-weight architecture of a 3×6 sized kernel can simultaneously implement parallel multiply-accumulate (MAC) operations for 144 bits inputs and recognize patterns up to 8 categories autonomously. Furthermore, it can connect with another DNA circuits to construct hierarchical networks, which can recognize patterns up to 32 categories with a two-step classification approach of performing coarse classification on language (Arabic numerals, Chinese oracles, English alphabets and Greek alphabets) and then classifying them into specific handwritten symbols. With a simple cyclic freeze/thaw approach, we can decrease computation time from hours to minutes. Our approach shows great promise in the realization of high computing power molecular computer with ability to classify complex and noisy information.

Information Content Measurement of ESG Factors via Entropy and Its Impact on Society and Security

We conducted a singular and sectoral vulnerability assessment of ESG factors of Dow-30-listed companies by applying the entropy weight method and analyzing each ESG factor’s information contribution to the overall ESG disclosure score. By reducing information entropy information, weaknesses in the structure of a socio-technological system can be identified and improved. The relative information gain of each indicator improves proportionally to the reduction in entropy. The social pillar contains the most crucial information, followed by the environmental and governance pillars, relative to each other. The difference between the social and economic pillars was found to be statistically not significant, while the differences between the social pillar, respective to the economic and governance pillars were statistically significant. This suggests noisy information content of the governance pillar, indicating improvement potential in governance messaging. Moreover, we found that companies with lean and flexible governance structures are more likely to convey information content better. We also discuss the impact of ESG measures on society and security.

A Nested Chinese Restaurant Topic Model for Short Texts with Document Embeddings

In recent years, short texts have become a kind of prevalent text on the internet. Due to the short length of each text, conventional topic models for short texts suffer from the sparsity of word co-occurrence information. Researchers have proposed different kinds of customized topic models for short texts by providing additional word co-occurrence information. However, these models cannot incorporate sufficient semantic word co-occurrence information and may bring additional noisy information. To address these issues, we propose a self-aggregated topic model incorporating document embeddings. Aggregating short texts into long documents according to document embeddings can provide sufficient word co-occurrence information and avoid incorporating non-semantic word co-occurrence information. However, document embeddings of short texts contain a lot of noisy information resulting from the sparsity of word co-occurrence information. So we discard noisy information by changing the document embeddings into global and local semantic information. The global semantic information is the similarity probability distribution on the entire dataset and the local semantic information is the distances of similar short texts. Then we adopt a nested Chinese restaurant process to incorporate these two kinds of information. Finally, we compare our model to several state-of-the-art models on four real-world short texts corpus. The experiment results show that our model achieves better performances in terms of topic coherence and classification accuracy.

Multi-Modal Residual Perceptron Network for Audio–Video Emotion Recognition

Emotion recognition is an important research field for human–computer interaction. Audio–video emotion recognition is now attacked with deep neural network modeling tools. In published papers, as a rule, the authors show only cases of the superiority in multi-modality over audio-only or video-only modality. However, there are cases of superiority in uni-modality that can be found. In our research, we hypothesize that for fuzzy categories of emotional events, the within-modal and inter-modal noisy information represented indirectly in the parameters of the modeling neural network impedes better performance in the existing late fusion and end-to-end multi-modal network training strategies. To take advantage of and overcome the deficiencies in both solutions, we define a multi-modal residual perceptron network which performs end-to-end learning from multi-modal network branches, generalizing better multi-modal feature representation. For the proposed multi-modal residual perceptron network and the novel time augmentation for streaming digital movies, the state-of-the-art average recognition rate was improved to 91.4% for the Ryerson Audio–Visual Database of Emotional Speech and Song dataset and to 83.15% for the Crowd-Sourced Emotional Multi Modal Actors dataset. Moreover, the multi-modal residual perceptron network concept shows its potential for multi-modal applications dealing with signal sources not only of optical and acoustical types.