@article{MTMT:31144197,
	title = {Parallel MCMC sampling of AR-HMMs for prediction based option trading},
	url = {https://m2.mtmt.hu/api/publication/31144197},
	author = {Sipos, István Róbert and Ceffer, Attila and Horváth, Gábor and Levendovszky, János},
	doi = {10.3233/AF-190243},
	journal-iso = {ALGORITHM FINANCE},
	journal = {ALGORITHMIC FINANCE},
	volume = {8},
	unique-id = {31144197},
	issn = {2158-5571},
	year = {2019},
	eissn = {2157-6203},
	pages = {47-55},
	orcid-numbers = {Horváth, Gábor/0000-0003-3097-1273; Levendovszky, János/0000-0003-1406-442X}
}

@article{MTMT:3252486,
	title = {Low Complexity Algorithmic Trading by Feedforward Neural Networks},
	url = {https://m2.mtmt.hu/api/publication/3252486},
	author = {Levendovszky, János and Reguly, István Zoltán and Oláh, András and Ceffer, Attila},
	doi = {10.1007/s10614-017-9720-6},
	journal-iso = {COMPUT ECON},
	journal = {COMPUTATIONAL ECONOMICS},
	volume = {54},
	unique-id = {3252486},
	issn = {0927-7099},
	year = {2019},
	eissn = {1572-9974},
	pages = {267-279},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X; Reguly, István Zoltán/0000-0002-4385-4204; Oláh, András/0009-0003-4796-8932}
}

@article{MTMT:3393080,
	title = {Trading by estimating the quantized forward distribution},
	url = {https://m2.mtmt.hu/api/publication/3393080},
	author = {Ceffer, Attila and Fogarasi, Norbert and Levendovszky, János},
	doi = {10.1080/00036846.2018.1486021},
	journal-iso = {APPL ECON},
	journal = {APPLIED ECONOMICS},
	volume = {50},
	unique-id = {3393080},
	issn = {0003-6846},
	year = {2018},
	eissn = {1466-4283},
	pages = {6397-6405},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X}
}

@article{MTMT:3256024,
	title = {Applying Independent Component Analysis and Predictive Systems for Algorithmic Trading},
	url = {https://m2.mtmt.hu/api/publication/3256024},
	author = {Ceffer, Attila and Levendovszky, János and Fogarasi, Norbert},
	doi = {10.1007/s10614-017-9719-z},
	journal-iso = {COMPUT ECON},
	journal = {COMPUTATIONAL ECONOMICS},
	volume = {54},
	unique-id = {3256024},
	issn = {0927-7099},
	year = {2018},
	eissn = {1572-9974},
	pages = {281-303},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X}
}

@CONFERENCE{MTMT:3342883,
	title = {Trading by estimating the forward distribution using quantization and volatility information},
	url = {https://m2.mtmt.hu/api/publication/3342883},
	author = {Ceffer, Attila and Levendovszky, János},
	booktitle = {3rd International Workshop on Financial Markets and Nonlinear Dynamics},
	unique-id = {3342883},
	year = {2017},
	pages = {1-14},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X}
}

@article{MTMT:3043217,
	title = {Parallel optimization of sparse portfolios with AR-HMMs},
	url = {https://m2.mtmt.hu/api/publication/3043217},
	author = {Sipos, István Róbert and Ceffer, Attila and Levendovszky, János},
	doi = {10.1007/s10614-016-9579-y},
	journal-iso = {COMPUT ECON},
	journal = {COMPUTATIONAL ECONOMICS},
	volume = {49},
	unique-id = {3043217},
	issn = {0927-7099},
	year = {2017},
	eissn = {1572-9974},
	pages = {563-578},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X}
}

@CONFERENCE{MTMT:3183235,
	title = {Applying ICA and NARX networks for algorithmic trading},
	url = {https://m2.mtmt.hu/api/publication/3183235},
	author = {Ceffer, Attila and Levendovszky, János and Fogarasi, Norbert},
	booktitle = {Fourth International Symposium in Computational Economics and Finance},
	unique-id = {3183235},
	abstract = {In this paper, a Nonlinear AutoRegressive network with 
		eXogenous inputs (NARX) is proposed for algorithmic trading by 
		predicting the future value of financial time series. This 
		network is highly capable of modeling vector autoregressive 
		VAR(p) time series. In order to avoid overfitting, the input 
		is pre-processed by Independent Component Analysis (ICA) to 
		filter out the most noise like component. In this way, the
		accuracy of the prediction is increased. The proposed 
		algorithm has a reduced number of free parameters which makes 
		fast learning and trading possible. The method is not only 
		tested on single asset price series, but also on predicting 
		the value of mean reverting portfolios obtained by maximizing 
		the predictability parameter of VAR(1) processes. The tests 
		were first performed on artificially generated data and then 
		on real data selected from Exchange Traded Fund (ETF) time 
		series including bid-ask spread. In both cases profit has been 
		achieved by the proposed method.},
	year = {2016},
	pages = {1-16}
}

@inproceedings{MTMT:3183215,
	title = {Kolmogorov-Smirnov test for keystroke dynamics based user authentication},
	url = {https://m2.mtmt.hu/api/publication/3183215},
	author = {Ceffer, Attila and Levendovszky, János},
	booktitle = {2016 17TH IEEE INTERNATIONAL SYMPOSIUM ON COMPUTATIONAL INTELLIGENCE AND INFORMATICS (CINTI 2016},
	doi = {10.1109/CINTI.2016.7846387},
	unique-id = {3183215},
	abstract = {In this paper a novel user authentication method is proposed 
		based on analyzing the keystroke patterns. Instead of using 
		Nearest Neighbor classification with Dynamic Time Warping 
		(DTW) distance measure, the typing dynamics are classified by 
		the Kolmogorov-Smirnov test. First the typing pattern is 
		translated into a sequence of holding times and latencies 
		between consecutive keystrokes. In order to increase the 
		classification performance, an adaptive preprocessing is used 
		to get rid of the outliers. Then the Kolmogorov-Smirnov test 
		is applied to classify the observed pattern. To further 
		increase the correct classification ratio, semi-supervised 
		self training methods are applied. One of the key objectives 
		of the paper is to analyze the performance with respect to the 
		length of typed characters, as in the case of authentication 
		systems users cannot be asked to type long texts. The 
		simulations have demonstrated that the proposed method 
		performs better than the 1 Nearest Neighbor DTW classifier on 
		all text lengths. It has also been shown that when the length 
		of the typed text reduces to 10 characters, then the 
		classification ratio sinks to 50% from 90.5% achieved in the 
		case of longer texts in the range of 200 characters.},
	year = {2016},
	pages = {105-110},
	orcid-numbers = {Levendovszky, János/0000-0003-1406-442X}
}