Sharif Digital Repository

Despite recent advancements in deep learning methods, these methods rely on a huge amount of training data to work. Recently the problem of solving classification and recently semantic segmentation problems with a few training data have gained attention to tackle this issue. In this research, we propose a meta-learning method by combining optimization-based and prototypical approaches in which a small portion of parameters are optimized with task-specific initialization. In addition to this and designing other parts of the method, we propose a new approach to use query data as an unlabeled sample to enhance task-specific learning. Alongside the mentioned method, we propose an approach to use... 

The goal of drug design is to identify new molecules with a set of desirable properties. The molecular search space is large, discrete, and unstructured, which results in a prolonged construction and testing process of new compounds and requires significant costs. Furthermore, there is a wide variety of appealing options to choose from. Recent advances in the field of machine learning have led to the emergence of generative models that, after training on real examples, can suggest suitable molecules with less time and cost. One of the stages that should be considered in the path of drug production is predicting the properties of the chemical molecule and its effect on the desired protein. By... 

Many real-world environment challenges are not considered in laboratory-controlled models. Although different and powerful models have been developed for object detection and classification in diverse applications, many fail in the real world. One of the most important challenges is dealing with unknown data at the inference time. The second challenge is to change the characteristics of the data distribution over time, known as concept drift. These two important challenges are explored in the Data Stream environment, along with many of the events that a model may face in the real world. To address the challenges of learning in a data stream environment, this thesis first designs a... 

Deep reinforcement learning has achieved better cumulative rewards than humans in many environments like Atari. One drawback of these methods is their data inefficiency which makes training time-consuming, and in some cases having this amount of data is infeasible. Meta reinforcement learning can use past experiences to enable agents to adapt to new tasks faster and makes neural networks to train in a short amount of time.One of the methods in meta reinforcement learning is inferring tasks which helps exploitation policy to have good performance in new tasks. There’s a need to improve exploration policy as well as exploitation policy by gaining informative transitions about the new task.... 

The problem of stock price direction prediction is of great value among investors and researchers in the past decades. Even the smallest improvement in the performance of forecasting methods can lead to noticeable profit for investors. In this regard, in this research, a new method for filling the literature gap in the field of stock price direction forecasting is proposed. In the proposed method, two concepts of dynamics and model selection in dealing with data is investigated. Finally a predictive model is developed according to the two abovementioned concepts. Moreover, in this work, using a meta-learning approach one step towards making the prediction process automatic is taken. The... 

This research aims to design and evaluate a generalized algorithm for predicting knee joint angle during gait. To achieve this, electromyography (EMG) sensors were attached to muscles involved in knee joint motion to record their activity. By analyzing and extracting features from these signals, a machine learning model was developed to establish a relationship between the EMG signals as input and the knee joint angle as output. The study prioritizes a model that can be quickly adapted to new users with minimal data, making meta-learning the core approach. This method shares similarities with transfer learning and was trained using two datasets, each containing data from over 10 subjects.... 

Nowadays, the capital market has a significant impact on the economy of a country and causes economic dynamism and growth in gross production. Among the important phenomena in the stock market is stock pricing, the correctness or incorrectness of which has a significant role in the performance of the stock market and the value of companies. The stock price in the stock exchange represents the stock market value and usually represents the investment value of the shareholders. Forecasting the trend of the stock market is considered an important and necessary thing and has been given much attention, because the successful forecasting of the stock price may lead to attractive profits by making... 

Few-shot segmentation (FSS) models have gained popularity in medical imaging analysis due to their ability to generalize well to unseen classes with only a small amount of annotated data. A key requirement for the success of FSS models is a diverse set of annotated classes as the base training tasks. This is a difficult condition to meet in the medical domain due to the lack of annotations, especially in volumetric images. To tackle this problem, self-supervised FSS methods for 3D images have been introduced. However, existing methods often ignore intravolume information in 3D image segmentation, which can limit their performance. To address this issue, we propose a novel selfsupervised... 

User cold-start is a common problem among real-world applications in the sequential recommendation field since determining user preference based on a few interactions is difficult. The problem would end up limiting the performance recommender systems. To address the cold-start problem, some previous works used meta-learning along with user’s and item’s side information. Meta-learning algorithms made the model able to share knowledge among all tasks. Although they had promising results, they had some fundamental issues with modeling the dynamics of user preferences and considering all kinds of users’ preferences, especially for minor users. The proposed method includes a model incorporating...