A hybrid supervised semi-supervised graph-based model to predict one-day ahead movement of global stock markets and commodity prices

Kia, Arash Negahdari; Haratizadeh, Saman; Shouraki, Saeed Bagheri

doi:10.1016/j.eswa.2018.03.037

Cited by 35 publications

(17 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the domain of time series prediction, semi-supervised algorithms are less widespread. Short-term prediction in the stock exchange market with help of semi-supervised algorithms is proposed in [12]. In our work, we do not transfer derived labels, as in semi-supervised learning, but similarity information from another domain, and consider this information for short-term and long-term predictions.…”

Section: Label Prediction Under Infinite Verification Latencymentioning

confidence: 99%

Entity-level stream classification: exploiting entity similarity to label the future observations referring to an entity

Unnikrishnan

Beyer

Matuszyk

et al. 2019

Int J Data Sci Anal

View full text Add to dashboard Cite

Stream classification algorithms traditionally treat arriving instances as independent. However, in many applications, the arriving examples may depend on the "entity" that generated them, e.g. in product reviews or in the interactions of users with an application server. In this study, we investigate the potential of this dependency by partitioning the original stream of instances/"observations" into entity-centric substreams and by incorporating entity-specific information into the learning model. We propose a k-nearest-neighbour-inspired stream classification approach, in which the label of an arriving observation is predicted by exploiting knowledge on the observations belonging to this entity and to entities similar to it. For the computation of entity similarity, we consider knowledge about the observations and knowledge about the entity, potentially from a domain/feature space different from that in which predictions are made. To distinguish between cases where this knowledge transfer is beneficial for stream classification and cases where the knowledge on the entities does not contribute to classifying the observations, we also propose a heuristic approach based on random sampling of substreams using k Random Entities (kRE). Our learning scenario is not fully supervised: after acquiring labels for the initial m observations of each entity, we assume that no additional labels arrive and attempt to predict the labels of near-future and far-future observations from that initial seed. We report on our findings from three datasets. Keywords Stream classification • kNN • Entity similarityWe use the terms observations and instances interchangeably, since we speak of a time series of property values that exist at all times but are observed at different points in time.

show abstract

Section: Label Prediction Under Infinite Verification Latencymentioning

confidence: 99%

Entity-level stream classification: exploiting entity similarity to label the future observations referring to an entity

Unnikrishnan

Beyer

Matuszyk

et al. 2019

Int J Data Sci Anal

View full text Add to dashboard Cite

show abstract

“…The DST shrinks significantly over time while applied to analyze the network of the stock indices from the Asia-Pacific [46]. MST can also be considered the starting point, and further edges can be added to that if the new edges can improve the performance of the subsequent task, such as direction prediction [27].…”

Section: Minimum Spanning Tree (Mst)-based Approachmentioning

confidence: 99%

“…It is possible to use a combination of supervised and semisupervised models while predicting stock movement using a graph-based approach [27,40]. ML models such as ANN or SVM use past time series data for predicting the future direction of different global indices in the supervised part.…”

Section: Traditional Statistical and Machine Learning Modelsmentioning

confidence: 99%

“…ML models such as ANN or SVM use past time series data for predicting the future direction of different global indices in the supervised part. The semi-supervised part uses the label spreading technique to predict the node type (e.g., rise or fall) for the next period [27]. The main idea is to use the already available label (rise or fall) of the other global markets in a semi-supervised manner to predict the direction of the unlabelled markets [27].…”

Section: Traditional Statistical and Machine Learning Modelsmentioning

confidence: 99%

See 1 more Smart Citation

A survey of the application of graph-based approaches in stock market analysis and prediction

Saha

Gao

Gerlach

2022

Int J Data Sci Anal

View full text Add to dashboard Cite

Graph-based approaches are revolutionizing the analysis of different real-life systems, and the stock market is no exception. Individual stocks and stock market indices are connected, and interesting patterns appear when the stock market is considered as a graph. Researchers are analyzing the stock market using graph-based approaches in recent years, and there is a need to survey those works from multiple perspectives. We discuss the existing graph-based works from five perspectives: (i) stock market graph formulation, (ii) stock market graph filtering, (iii) stock market graph clustering, (iv) stock movement prediction, and (v) portfolio optimization. This study contains a concise description of major techniques and algorithms relevant to graph-based approaches for the stock market.

show abstract

“…However, for transforming a time series into a signal series, the selected thresholds should be partly subjective and it is easy to make different similarity matrixes by choosing different thresholds, since they are not chosen totally objective. The last approach is based on a correlation coefficient matrix, which itself is a similarity matrix [18], [19]. Due to its convenient computation, the correlation coefficient matrix is the most popular and widely used similarity measurement among these three methods.…”

Section: Introductionmentioning

confidence: 99%

Projecting Financial Technical Indicators Into Networks as a Tool to Build a Portfolio

Chen

2021

IEEE Access

View full text Add to dashboard Cite

Using the topological structure of financial networks to build a portfolio has attracted a wide range of research interests. A similarity matrix based on the technical indicators (TIs), and a correlation matrix based on the stock returns, are used to construct the financial networks. Hybrid topological measures are calculated for both networks to select the stocks that are further used to build portfolios. The random matrix theory (RMT) tool is used to filter the risk measurement in the Markowitz optimization process. A general result is presented in this study, which indicates that the portfolios composed of peripheral stocks consistently outperform those composed of central stocks. As shown in the empirical results, the RMT method can improve the Markowitz optimization process and further improve the performance of portfolios. By comparing the different portfolios established under different networks, the results reveal the advantage of the topological measure based on TIs. This study provides an important approach for constructing financial networks from the time series data, and it explores the impact of different similarity measurements on the network-based models for portfolio selection.

show abstract

A hybrid supervised semi-supervised graph-based model to predict one-day ahead movement of global stock markets and commodity prices

Cited by 35 publications

References 33 publications

Entity-level stream classification: exploiting entity similarity to label the future observations referring to an entity

Entity-level stream classification: exploiting entity similarity to label the future observations referring to an entity

A survey of the application of graph-based approaches in stock market analysis and prediction

Projecting Financial Technical Indicators Into Networks as a Tool to Build a Portfolio

Contact Info

Product

Resources

About