Abstract: Disclosed herein is a denoising device including a deriving part configured to, when corrupted noise data corrupted due to noises is received from source data, derive an estimated loss which is estimated when each symbol within noise data is reconstructed to the source data based on a predefined noise occurrence probability, a processor to process training of a defined learning model by including parameters related with the reconstruction of the source data from the noise data based on context composed of a sequence of neighbored symbols based on each symbol within the noise data and pseudo-training data using the estimated loss corresponding to the context, and an output part to output reconstructed data in which each symbol within the noise data is reconstructed to a symbol of the source data through a denoiser formed based on a result of the training processing.

Abstract: A neural network training method based on training data, includes receiving training data including sequential data, and selecting a reference hidden node from hidden nodes in a neural network. The method further includes training the neural network based on remaining hidden nodes obtained by excluding the reference hidden node from the hidden nodes, and based on the training data, the remaining hidden nodes being connected with hidden nodes in a different time interval, and a connection between the reference hidden node and the hidden nodes in the different time interval being ignored.

Abstract: A denoising device, and a noise removal method. The denoising device includes a deriving part configured to, when corrupted noise data corrupted due to noises is received from source data, derive an estimated loss which is estimated when each symbol within noise data is reconstructed to the source data based on a predefined noise occurrence probability, a processor configured to process training of a defined learning model by including parameters related with the reconstruction of the source data from the noise data based on context composed of a sequence of neighbored symbols based on each symbol within the noise data and pseudo-training data using the estimated loss corresponding to the context, and an output part configured to output reconstructed data in which each symbol within the noise data is reconstructed to a symbol of the source data through a denoiser formed based on a result of the training processing.

Abstract: Disclosed is a neural network training method and apparatus, and recognition method and apparatus. The neural network training apparatus receives data and train a neural network based on remaining hidden nodes obtained by excluding a reference hidden node from hidden nodes included in the neural network, wherein the reference hidden node maintains a value in a previous time interval until a subsequent time interval.

Abstract: A neural network training method based on training data, includes receiving training data including sequential data, and selecting a reference hidden node from hidden nodes in a neural network. The method further includes training the neural network based on remaining hidden nodes obtained by excluding the reference hidden node from the hidden nodes, and based on the training data, the remaining hidden nodes being connected with hidden nodes in a different time interval, and a connection between the reference hidden node and the hidden nodes in the different time interval being ignored.

Abstract: Methods, systems, and computer programs are presented for providing internet content, such as related news articles. One method includes an operation for defining a plurality of candidates based on a seed. For each candidate, scores are calculated for relevance, novelty, connection clarity, and transition smoothness. The score for connection clarity is based on a relevance score of the intersection between the words in the seed and the words in each of the candidates. Further, the score for transition smoothness measures the interest in reading each candidate when transitioning from the seed to the candidate. For each candidate, a relatedness score is calculated based on the calculated scores for relevance, novelty, connection clarity, and transition smoothness. In addition, at least one of the candidates is selected based on their relatedness scores for presentation to the user.

Abstract: Methods, systems, and computer programs are presented for providing internet content, such as related news articles. One method includes an operation for defining a plurality of candidates based on a seed. For each candidate, scores are calculated for relevance, novelty, connection clarity, and transition smoothness. The score for connection clarity is based on a relevance score of the intersection between the words in the seed and the words in each of the candidates. Further, the score for transition smoothness measures the interest in reading each candidate when transitioning from the seed to the candidate. For each candidate, a relatedness score is calculated based on the calculated scores for relevance, novelty, connection clarity, and transition smoothness. In addition, at least one of the candidates is selected based on their relatedness scores for presentation to the user.