Hadi Sadoghi Yazdi

The Adaptive Neuro-Fuzzy Inference System (ANFIS) is a regression model that combines fuzzy logic and neural networks, making it suitable for modeling the uncertainty of regression problems. However, the non-robust loss function in ANFIS’s hybrid learning algorithm can render it vulnerable to the direct effects of noise and outliers. This paper introduces a new procedure that utilizes robust loss functions to enhance the hybrid learning performance against noise and outliers. Furthermore, a new robust loss function is developed that can completely ignore outliers. A set of robust loss functions with mathematical relations are also suggested. The proposed approach is evaluated on real-world problems, including weather forecasting and stock market prediction. Results indicate that the proposed model can reduce the Mean Square Error (MSE) in regression.

Semantic labeling for image datasets is of significant importance in a wide range of social media. However, social datasets with massive amounts of data require effective technologies to increase the quality of classification. In this study, we propose a novel online robust classification using distributed learning method, in which the diffusion method is used over adaptive networks in order to the parallelization of the training process. The loss function of the suggested method is derived from the logarithm cosine function, being more suitable for impulsive noises. Also, the convergence of the proposed method is discussed theoretically. The Extensive experimental results show that the proposed method outperforms the other robust state-of-the-art classification algorithms in the presence of various scenarios, such as: free noise synthetic data sets, pure Gaussian noise, mixture of two Gaussian noises, impulse noise, and Alpha–Beta noise which are added to the synthetic data sets. In addition, the experiments were repeated for two different types of real data sets with real noises and outliers, i.e., UCI and 500PX social media data sets. The obtained results over UCI and 500px data sets on social media with real outliers and mislabeled samples confirmed the acceptable performance of the proposed method.

The present study utilizes the social media data to sample from people’s emotion in different places based on their facial expressions, for analyzing emotions distribution in a city and comparing the emotional similarity between cities. Experiencing various emotions g1ives birth to different facial expressions which usually have similar patterns for individuals. Extracting these patterns from face images and analyzing them makes it possible to extract related emotion associated with the shared photos. Since geo-tagged images include metadata about the geographical location, estimation of emotions spatial distribution is possible. The distribution of four emotions: anger, disgust, happiness, and surprise in twelve cities, included Athens, Beijing, Berlin, Brussels, Buenos Aires, Copenhagen, Helsinki, Melbourn, New York, Ottawa, Paris, and Prague is investigated by analyzing geo-tagged photos shared on Flickr. After facial expressions extraction, the spatial distribution functions of the emotions are calculated by kernel density estimation method. Emotional maps for cities are created using extracted features of Fourier transform spectrum of the spatial distributions which are robust against rotation and translation. Afterward, similar cities based on their emotional structure are clustered into appropriate groups. The results are beneficial for urban planners and social researchers to analyze the effect of the environment on peoples’ emotions. Furthermore, this could guide them to make right policies to improve the quality of life. Considering the four studied emtions, New York has the highest ratio in anger, Ottawa has the highest proportion of happiness and Copenhagen has the highest ratio in both surprise and disgust among the studied cities. In addition, an attempt is made to calculate the percentage of different emotions and dispersion of them as important emotional factors in understanding the emotional structure of cities. To measure the dispersion of happiness the coefficient of variation is computed. The results show that happiness dispersion in Berlin has a more uniform distribution according to the collected data. It seems both coefficients of variation and emotions percentage may be jointly effective in urban planning and cities comparison based on people emotional mode.