-
Notifications
You must be signed in to change notification settings - Fork 0
/
research.html
198 lines (154 loc) · 10.1 KB
/
research.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
<!DOCTYPE HTML>
<!--
Alpha by HTML5 UP
html5up.net | @ajlkn
Free for personal and commercial use under the CCA 3.0 license (html5up.net/license)
-->
<html>
<head>
<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-75K7S6L1ML"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-75K7S6L1ML');
</script>
<title>Kleanthis Malialis | Research areas</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
<link rel="stylesheet" href="assets/css/main.css" />
</head>
<body class="is-preload">
<div id="page-wrapper">
<!-- Header -->
<header id="header">
<h1><a href="index.html">Kleanthis Malialis</a></h1>
<nav id="nav">
<ul>
<li><a href="index.html">Home</a></li>
<li>
<a href="#" class="icon solid fa-angle-down">Research</a>
<ul>
<li><a href="projects.html">Projects</a></li>
<li><a href="research.html">Areas</a></li>
</ul>
</li>
<li><a href="publications.html">Publications</a></li>
<li><a href="teaching.html">Teaching</a></li>
<li><a href="contact.html">Contact</a></li>
</ul>
</nav>
</header>
<!-- Main -->
<section id="main" class="container">
<header>
<h2>Research areas</h2>
<!-- <p>Just an assorted selection of elements.</p> -->
</header>
<div class="row">
<div class="col-12">
<!-- Table -->
<section class="box">
<p>My research interests lie within the areas of <strong>learning from nonstationary, limited-labelled, and imbalanced data streams</strong> with applications in monitoring of critical infrastructures and in healthcare.</p>
<p>I'm also interested in <strong>reinforcement learning</strong> and <strong>multiagent coordination</strong> with applications in security (network intrusion detection and response, downing rogue drones).</p>
<p>Below is a list of areas that I have worked or have been working in, along with a short description, representative publications and code.</p>
</section>
</div>
</div>
<div class="row">
<div class="col-6 col-12-narrower">
<section class="box special">
<span class="image featured"><img src="images/pic02.jpg" alt="" /></span>
<h3>Learning from imbalanced data streams</h3>
<p><img src="my_images/proj_areba.png" alt="" width="420" height="180"/></p>
<p>Learning in nonstationary environments constitutes a major challenge, and this problem becomes orders of magnitude more complex in the presence of class imbalance. Learning from nonstationary and imbalanced data has been studied separately, but many key challenges remain open when the joint problem is considered.</p>
<p><strong>Representative publications</strong>:
IEEE TNNLS 2020 <a href="https://ieeexplore.ieee.org/document/9203853">[pdf]</a> <a href="https://github.com/kmalialis/areba/">[code]</a>,
IEEE SSCI 2022 <a href="https://arxiv.org/abs/2210.04949">[pdf]</a>,
ICANN 2018 <a href="https://arxiv.org/abs/1809.10388">[pdf]</a> <a href="https://github.com/kmalialis/queue_based_resampling">[code]</a>
</p>
<!-- <ul class="actions special">
<li><a href="#" class="button alt">Learn More</a></li>
</ul> -->
</section>
</div>
<div class="col-6 col-12-narrower">
<section class="box special">
<span class="image featured"><img src="images/pic03.jpg" alt="" /></span>
<h3>Active learning from data streams</h3>
<p><img src="my_images/proj_actisiamese.png" alt="" width="370" height="200"/></p>
<p>Apart from data nonstationarity (concept drift) and class imbalance, acquiring ground truth information (e.g., labels in classification tasks) as instances arrive one-by-one can be costly or impossible in some real-time applications. An effective way to deal with limited labelled data is the active learning paradigm.</p>
<p><strong>Representative publications</strong>:
Neurocomputing 2022 <a href="https://arxiv.org/abs/2210.01090">[pdf]</a> <a href="https://github.com/kmalialis/actisiamese">[code]</a>,
IEEE WCCI (IJCNN) 2020 <a href="https://arxiv.org/abs/2010.01659">[pdf]</a>,
IEEE SSCI 2022 <a href="https://arxiv.org/abs/2210.06873">[pdf]</a> <a href="https://github.com/kmalialis/augmented_queues/">[code]</a>
</p>
</section>
</div>
<div class="col-6 col-12-narrower">
<section class="box special">
<span class="image featured"><img src="images/pic03.jpg" alt="" /></span>
<h3>Anomaly detection in data streams</h3>
<p><img src="my_images/proj_anomdet.png" alt="" width="350" height="380"/></p>
<p>The generation of vast amounts of streaming data in various domains has become ubiquitous. However, many of these data are unlabeled, making it challenging to identify infrequent events, particularly anomalies. This task becomes even more formidable in nonstationary environments where model performance can deteriorate over time due to concept drift. To address these challenges, we propose autoencoder-based incremental learning and concept drift detection mechanisms.</p>
<p><strong>Representative publications</strong>:
IEEE WCCI (IJCNN) 2024 <a href="https://arxiv.org/abs/2403.03576">[pdf]</a>,
IEEE IJCNN 2023 <a href="https://arxiv.org/abs/2305.08977">[pdf]</a>,
IEEE SSCI 2023 <a href="https://arxiv.org/abs/2211.12989">[pdf]</a>
</p>
</section>
</div>
<div class="col-6 col-12-narrower">
<section class="box special">
<span class="image featured"><img src="images/pic03.jpg" alt="" /></span>
<h3>Multiagent reinforcement learning for intrusion detection and response</h3>
<p><img src="my_images/proj_marl.png" alt="" width="300" height="190"/></p>
<p>A serious threat in the current Internet is distributed denial of service (DDoS) attacks, which target the availability of the victim system. They are designed to exhaust a server's resources or congest a network's infrastructure, and therefore renders the victim incapable of providing services to its legitimate users. To address this, a distributed and coordinated defence mechanism is necessary, where many defensive nodes, across different locations cooperate in order to stop or reduce the flood. We propose the use of multiagent reinforcement learning to address the problem.</p>
<p><strong>Representative publications</strong>:
EAAI 2015 <a href="https://www.sciencedirect.com/science/article/abs/pii/S095219761500024X">[pdf]</a> (<span style="color:red">Department's Best Student Paper 2015 Award</span>),
Connection Science 2015 <a href="https://www.tandfonline.com/doi/full/10.1080/09540091.2015.1031082">[pdf]</a>,
AAAI / IAAI 2013 <a href="https://cdn.aaai.org/ojs/19000/19000-13-22812-1-10-20211006.pdf">[pdf]</a>,
AAMAS 2016 <a href="https://arxiv.org/abs/1903.05431">[pdf]</a> <a href="https://github.com/kmalialis/resource_abstraction/">[code]</a>,
EUMAS 2013 <a href="https://www.irit.fr/EUMAS2013/Papers/eumas2013_submission_21.pdf">[pdf]</a>,
PhD thesis <a href="http://etheses.whiterose.ac.uk/8109/">[pdf]</a>
</p>
</section>
</div>
<div class="col-6 col-12-narrower">
<section class="box special">
<span class="image featured"><img src="images/pic03.jpg" alt="" /></span>
<h3>Multiagent reinforcement learning for downing rogue drones</h3>
<p><img src="my_images/proj_drones.png" alt="" width="350" height="240"/></p>
<p>The wide adoption and use of unmanned aerial vehicles (UAVs) has created not only opportunities but also threats to the security of sensitive areas. Thus, effective and efficient counter-drone systems are required to protect these areas. This work addresses this issue by developing cooperative multi-agent searching, tracking and jamming techniques using RL to counter the operation of one or multiple rogue drones flying over a sensitive area.</p>
<p><strong>Representative publications</strong>:
IEEE TMC 2024 (PDF link TBA), IEEE SMC 2024 (PDF link TBA), ICUAS 2023 (PDF link TBA)
</p>
</section>
</div>
</div>
</section>
<!-- Footer -->
<footer id="footer">
<ul class="icons">
<li><a href="https://cy.linkedin.com/in/kleanthis-malialis-phd-0542a911a"><span class="label"><img src="my_images/logo_linkedin.png" alt="LinkedIn's logo" style="width:35px; height: 30px"></span></a></li>
<li><a href="https://twitter.com/kmalialis?lang=en"><span class="label"><img src="my_images/logo_x.png" alt="X's logo" style="width:30px; height: 30px"></span></a></li>
<li><a href="https://scholar.google.com/citations?user=O2Zqu2sAAAAJ&hl=en"><span class="label"><img src="my_images/logo_gscholar.png" alt="Google Scholar's logo" style="width:32px; height: 32px"></span></a></li>
<li><a href="https://github.com/kmalialis" class="icon brands fa-github"><span class="label">Github</span></a></li>
<li><a href="https://orcid.org/0000-0003-3432-7434"><span class="label"><img src="my_images/logo_orcid.png" alt="ORCiD's logo" style="width:30px; height: 30px"></span></a></li>
</ul>
<ul class="copyright">
<li>© Kleanthis Malialis. All rights reserved.</li>
</ul>
</footer>
</div>
<!-- Scripts -->
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/jquery.dropotron.min.js"></script>
<script src="assets/js/jquery.scrollex.min.js"></script>
<script src="assets/js/browser.min.js"></script>
<script src="assets/js/breakpoints.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>