his.sePublications
Change search
Link to record
Permanent link

Direct link
BETA
Publications (10 of 26) Show all publications
Ståhl, N., Falkman, G., Karlsson, A., Mathiason, G. & Boström, J. (2019). Deep Reinforcement Learning for Multiparameter Optimization in de novo Drug Design. Journal of Chemical Information and Modeling, 59(7), 3166-3176
Open this publication in new window or tab >>Deep Reinforcement Learning for Multiparameter Optimization in de novo Drug Design
Show others...
2019 (English)In: Journal of Chemical Information and Modeling, ISSN 1549-9596, E-ISSN 1549-960X, Vol. 59, no 7, p. 3166-3176Article in journal (Refereed) Published
Abstract [en]

In medicinal chemistry programs it is key to design and make compounds that are efficacious and safe. This is a long, complex, and difficult multiparameter optimization process, often including several properties with orthogonal trends. New methods for the automated design of compounds against profiles of multiple properties are thus of great value. Here we present a fragment-based reinforcement learning approach based on an actor-critic model, for the generation of novel molecules with optimal properties. The actor and the critic are both modeled with bidirectional long short-term memory (LSTM) networks. The AI method learns how to generate new compounds with desired properties by starting from an initial set of lead molecules and then improving these by replacing some of their fragments. A balanced binary tree based on the similarity of fragments is used in the generative process to bias the output toward structurally similar molecules. The method is demonstrated by a case study showing that 93% of the generated molecules are chemically valid and more than a third satisfy the targeted objectives, while there were none in the initial set.

Keywords
algorithms, molecules
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-17503 (URN)10.1021/acs.jcim.9b00325 (DOI)000477074900010 ()31273995 (PubMedID)2-s2.0-85070180995 (Scopus ID)
Available from: 2019-08-08 Created: 2019-08-08 Last updated: 2019-11-08Bibliographically approved
Ståhl, N., Falkman, G., Karlsson, A., Mathiason, G. & Boström, J. (2019). Improving the use of deep convolutional neural networks for the prediction of molecular properties. In: Florentino Fdez-Riverola, Mohd Saberi Mohamad, Miguel Rocha, Juan F. De Paz, Pascual González (Ed.), Practical Applications of Computational Biology and Bioinformatics, 12th International Conference: . Paper presented at PACBB2018: International Conference on Practical Applications of Computational Biology & Bioinformatics, Toledo, June 20-22, 2018 (pp. 71-79). Cham: Springer, 803
Open this publication in new window or tab >>Improving the use of deep convolutional neural networks for the prediction of molecular properties
Show others...
2019 (English)In: Practical Applications of Computational Biology and Bioinformatics, 12th International Conference / [ed] Florentino Fdez-Riverola, Mohd Saberi Mohamad, Miguel Rocha, Juan F. De Paz, Pascual González, Cham: Springer, 2019, Vol. 803, p. 71-79Conference paper, Published paper (Refereed)
Abstract [en]

We present a flexible deep convolutional neural network method for the analyse of arbitrary sized graph structures representing molecules. The method makes use of RDKit, an open-source cheminformatics software, allowing the incorporation of any global molecular (such as molecular charge) and local (such as atom type) information. We evaluate the method on the Side Effect Resource (SIDER) v4.1 dataset and show that it significantly outperforms another recently proposed method based on deep convolutional neural networks. We also reflect on how different types of information and input data affect the predictive power of our model. This reflection highlights several open problems that should be solved to further improve the use of deep learning within cheminformatics.

Place, publisher, year, edition, pages
Cham: Springer, 2019
Series
Advances in Intelligent Systems and Computing, ISSN 2194-5357, E-ISSN 2194-5365 ; 803
Keywords
drug discovery, graph convolutional neural network, molecular property prediction, bioinformatics, convolution, neural networks, open source software, open systems, cheminformatics, convolutional neural network, deep convolutional neural networks, graph structures, molecular charge, molecular properties, predictive power, deep neural networks
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL); INF301 Data Science
Identifiers
urn:nbn:se:his:diva-16230 (URN)10.1007/978-3-319-98702-6_9 (DOI)000468071900009 ()2-s2.0-85052956812 (Scopus ID)978-3-319-98701-9 (ISBN)978-3-319-98702-6 (ISBN)
Conference
PACBB2018: International Conference on Practical Applications of Computational Biology & Bioinformatics, Toledo, June 20-22, 2018
Available from: 2018-09-25 Created: 2018-09-25 Last updated: 2019-06-07Bibliographically approved
Bae, J., Havsol, J., Karpefors, M., Karlsson, A. & Mathiason, G. (2019). Short Text Topic Modeling to Identify Trends on Wearable Bio-sensors in Different Media Types. In: Proceedings - 6th International Symposium on Computational and Business Intelligence, ISCBI 2018: . Paper presented at ISCBI 2018 : 2018 6th International Symposium on Computational and Business Intelligence. Basel, Switzerland August 22 - 29 2018 (pp. 89-93). IEEE Computer Society
Open this publication in new window or tab >>Short Text Topic Modeling to Identify Trends on Wearable Bio-sensors in Different Media Types
Show others...
2019 (English)In: Proceedings - 6th International Symposium on Computational and Business Intelligence, ISCBI 2018, IEEE Computer Society, 2019, p. 89-93Conference paper, Published paper (Refereed)
Abstract [en]

The technology and techniques for bio-sensors are rapidly evolving. Accordingly, there is significant business interest to identify upcoming technologies and new targets for the near future. Text information from internet reflects much of the recent information and public interests that help to understand the trend of a certain field. Thus, we utilize Dirichlet process topic modeling on different media sources containing short text (e.g., blogs, news) which is able to self-adapt the learned topic space to the data. We share the observations from the domain experts on the results derived from topic modeling on wearable biosensors from multiple media sources over more than eight years. We analyze the topics on wearable devices, forecast and market analysis, and bio-sensing techniques found from our method. 

Place, publisher, year, edition, pages
IEEE Computer Society, 2019
Keywords
Bayesian non-parametrics, Bio-sensor, short text, topic modeling, wearable, Biosensors, Information analysis, Bayesian nonparametrics, Dirichlet process, Market analysis, Short texts, Text information, Wearable devices, Wearable sensors
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-16746 (URN)10.1109/ISCBI.2018.00027 (DOI)000462379700017 ()2-s2.0-85063041846 (Scopus ID)978-1-5386-9450-3 (ISBN)978-1-5386-9451-0 (ISBN)
Conference
ISCBI 2018 : 2018 6th International Symposium on Computational and Business Intelligence. Basel, Switzerland August 22 - 29 2018
Available from: 2019-04-05 Created: 2019-04-05 Last updated: 2019-09-30Bibliographically approved
Steinhauer, H. J., Helldin, T., Mathiason, G. & Karlsson, A. (2019). Topic modeling for anomaly detection in telecommunication networks. Journal of Ambient Intelligence and Humanized Computing, 1-12
Open this publication in new window or tab >>Topic modeling for anomaly detection in telecommunication networks
2019 (English)In: Journal of Ambient Intelligence and Humanized Computing, ISSN 1868-5137, E-ISSN 1868-5145, p. 1-12Article in journal (Refereed) Epub ahead of print
Abstract [en]

To ensure reliable network performance, anomaly detection is an important part of the telecommunication operators’ work. This includes that operators need to timely intervene with the network, should they encounter indications of network performance degradation. In this paper, we describe the results of an initial experiment for anomaly detection with regard to network performance, using topic modeling on base station run-time variable data collected from live Radio Access Networks (RANs). The results show that topic modeling clusters semantically related data in the same way as human experts would and that the anomalies in our test cases could be identified in latent Dirichlet allocation (LDA) topic models. Our experiment further reveals which information provided by the topic model is particularly usable to support human anomaly detection in this application domain.

Place, publisher, year, edition, pages
Springer, 2019
Keywords
Telecommunication anomaly detection, Topic modeling, Decision-making
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-17527 (URN)10.1007/s12652-019-01372-5 (DOI)
Available from: 2019-08-13 Created: 2019-08-13 Last updated: 2019-11-08Bibliographically approved
Ståhl, N., Mathiason, G., Falkman, G. & Karlsson, A. (2019). Using recurrent neural networks with attention for detecting problematic slab shapes in steel rolling. Applied Mathematical Modelling, 70, 365-377
Open this publication in new window or tab >>Using recurrent neural networks with attention for detecting problematic slab shapes in steel rolling
2019 (English)In: Applied Mathematical Modelling, ISSN 0307-904X, E-ISSN 1872-8480, Vol. 70, p. 365-377Article in journal (Refereed) Published
Abstract [en]

The competitiveness in the manufacturing industry raises demands for using recent data analysis algorithms for manufacturing process development. Data-driven analysis enables extraction of novel knowledge from already existing sensors and data, which is necessary for advanced manufacturing process refinement involving aged machinery. Improved data analysis enables factories to stay competitive against newer factories, but without any hefty investment. In large manufacturing operations, the dependencies between data are highly complex and therefore very difficult to analyse manually. This paper applies a deep learning approach, using a recurrent neural network with long short term memory cells together with an attention mechanism to model the dependencies between the measured product shape, as measured before the most critical manufacturing operation, and the final product quality. Our approach predicts the ratio of flawed products already before the critical operation with an AUC-ROC score of 0.85, i.e., we can detect more than 80 % of all flawed products while having less than 25 % false positive predictions (false alarms). In contrast to previous deep learning approaches, our method shows how the recurrent neural network reasons about the input shape, using the attention mechanism to point out which parts of the product shape that have the highest influence on the predictions. Such information is crucial for both process developers, in order to understand and improve the process, and for process operators who can use the information to learn how to better trust the predictions and control the process.

Place, publisher, year, edition, pages
Elsevier, 2019
Keywords
Attention mechanism, Recurrent Neural Networks, Interpretable AI, Steel Rolling
National Category
Computer Sciences
Research subject
INF301 Data Science; Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-16588 (URN)10.1016/j.apm.2019.01.027 (DOI)000468714000021 ()2-s2.0-85060941525 (Scopus ID)
Projects
DataFlow
Funder
Vinnova, 2017-01531
Available from: 2019-01-29 Created: 2019-01-29 Last updated: 2019-07-10Bibliographically approved
Ståhl, N., Falkman, G., Mathiason, G. & Karlsson, A. (2018). A self-organizing ensemble of deep neural networks for the classification of data from complex processes. In: Medina, J., Ojeda-Aciego, M., Verdegay, J.L., Perfilieva, I., Bouchon-Meunier, B., Yager, R.R. (Ed.), INFORMATION PROCESSING AND MANAGEMENT OF UNCERTAINTY IN KNOWLEDGE-BASED SYSTEMS: APPLICATIONS, IPMU 2018, PT III: . Paper presented at 17th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems, Jun 11 - Jun 15, 2018, Cadiz, Spain (pp. 248-259). , 855
Open this publication in new window or tab >>A self-organizing ensemble of deep neural networks for the classification of data from complex processes
2018 (English)In: INFORMATION PROCESSING AND MANAGEMENT OF UNCERTAINTY IN KNOWLEDGE-BASED SYSTEMS: APPLICATIONS, IPMU 2018, PT III / [ed] Medina, J., Ojeda-Aciego, M., Verdegay, J.L., Perfilieva, I., Bouchon-Meunier, B., Yager, R.R., 2018, Vol. 855, p. 248-259Conference paper, Published paper (Refereed)
Abstract [en]

We present a new self-organizing algorithm for classification of a data that combines and extends the strengths of several common machine learning algorithms, such as algorithms in self-organizing neural networks, ensemble methods and deep neural networks. The increased expression power is combined with the explanation power of self-organizing networks. Our algorithm outperforms both deep neural networks and ensembles of deep neural networks. For our evaluation case, we use production monitoring data from a complex steel manufacturing process, where data is both high-dimensional and has many nonlinear interdependencies. In addition to the improved prediction score, the algorithm offers a new deep-learning based approach for how computational resources can be focused in data exploration, since the algorithm points out areas of the input space that are more challenging to learn.

Series
Communications in Computer and Information Science, ISSN 1865-0929, E-ISSN 1865-0937
Keywords
artificial neural networks, complex processes, ensemble methods, self organisation
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL); Skövde Artificial Intelligence Lab (SAIL); INF301 Data Science
Identifiers
urn:nbn:se:his:diva-15008 (URN)10.1007/978-3-319-91479-4_21 (DOI)000481660700021 ()2-s2.0-85048061876 (Scopus ID)978-3-319-91479-4 (ISBN)978-3-319-91478-7 (ISBN)
Conference
17th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems, Jun 11 - Jun 15, 2018, Cadiz, Spain
Available from: 2018-04-04 Created: 2018-04-04 Last updated: 2019-09-05Bibliographically approved
Steinhauer, H. J., Helldin, T., Mathiason, G. & Karlsson, A. (2018). Anomaly Detection in Telecommunication Networks using Topic Models. In: : . Paper presented at Modeling Decisions for Artificial Intelligence, 15th International Conference, MDAI 2018, Mallorca, Spain, October 15–18, 2018.
Open this publication in new window or tab >>Anomaly Detection in Telecommunication Networks using Topic Models
2018 (English)Conference paper, Published paper (Refereed)
National Category
Computer and Information Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL); INF301 Data Science
Identifiers
urn:nbn:se:his:diva-16491 (URN)
Conference
Modeling Decisions for Artificial Intelligence, 15th International Conference, MDAI 2018, Mallorca, Spain, October 15–18, 2018
Available from: 2018-12-12 Created: 2018-12-12 Last updated: 2019-06-13Bibliographically approved
Ståhl, N., Falkman, G., Karlsson, A., Mathiason, G. & Boström, J. (2018). Deep Convolutional Neural Networks for the Prediction of Molecular Properties: Challenges and Opportunities Connected to the Data. Journal of Integrative Bioinformatics, 16(1)
Open this publication in new window or tab >>Deep Convolutional Neural Networks for the Prediction of Molecular Properties: Challenges and Opportunities Connected to the Data
Show others...
2018 (English)In: Journal of Integrative Bioinformatics, E-ISSN 1613-4516, Vol. 16, no 1Article in journal (Refereed) Published
Abstract [en]

We present a flexible deep convolutional neural network method for the analysis of arbitrary sized graph structures representing molecules. This method, which makes use of the Lipinski RDKit module, an open-source cheminformatics software, enables the incorporation of any global molecular (such as molecular charge and molecular weight) and local (such as atom hybridization and bond orders) information. In this paper, we show that this method significantly outperforms another recently proposed method based on deep convolutional neural networks on several datasets that are studied. Several best practices for training deep convolutional neural networks on chemical datasets are also highlighted within the article, such as how to select the information to be included in the model, how to prevent overfitting and how unbalanced classes in the data can be handled.

Place, publisher, year, edition, pages
De Gruyter Open, 2018
Keywords
Molecular property prediction, Deep learning, Unbalanced data, Side effects prediction
National Category
Computer Sciences Medicinal Chemistry
Research subject
Skövde Artificial Intelligence Lab (SAIL); INF301 Data Science
Identifiers
urn:nbn:se:his:diva-16480 (URN)10.1515/jib-2018-0065 (DOI)000459384300004 ()30517077 (PubMedID)2-s2.0-85061622864 (Scopus ID)
Available from: 2018-12-09 Created: 2018-12-09 Last updated: 2019-06-13Bibliographically approved
Karlsson, A., Duarte, D., Mathiason, G. & Bae, J. (2018). Evaluation of the dirichlet process multinomial mixture model for short-text topic modeling. In: Proceedings - 6th International Symposium on Computational and Business Intelligence, ISCBI 2018: . Paper presented at 6th International Symposium on Computational and Business Intelligence (ISCBI), 27-29 August 2018, Basel, Switzerland (pp. 79-83). USA: Institute of Electrical and Electronics Engineers (IEEE), Article ID 8638311.
Open this publication in new window or tab >>Evaluation of the dirichlet process multinomial mixture model for short-text topic modeling
2018 (English)In: Proceedings - 6th International Symposium on Computational and Business Intelligence, ISCBI 2018, USA: Institute of Electrical and Electronics Engineers (IEEE), 2018, p. 79-83, article id 8638311Conference paper, Published paper (Refereed)
Abstract [en]

Fast-moving trends, both in society and in highly competitive business areas, call for effective methods for automatic analysis. The availability of fast-moving sources in the form of short texts, such as social media and blogs, allows aggregation from a vast number of text sources, for an up to date view of trends and business insights. Topic modeling is established as an approach for analysis of large amounts of texts, but the scarcity of statistical information in short texts is considered to be a major problem for obtaining reliable topics from traditional models such as LDA. A range of different specialized topic models have been proposed, but a majority of these approaches rely on rather strong parametric assumptions, such as setting a fixed number of topics. In contrast, recent advances in the field of Bayesian non-parametrics suggest the Dirichlet process as a method that, given certain hyper-parameters, can self-adapt to the number of topics of the data at hand. We perform an empirical evaluation of the Dirichlet process multinomial (unigram) mixture model against several parametric topic models, initialized with different number of topics. The resulting models are evaluated, using both direct and indirect measures that have been found to correlate well with human topic rankings. We show that the Dirichlet Process Multinomial Mixture model is a viable option for short text topic modeling since it on average performs better, or nearly as good, compared to the parametric alternatives, while reducing parameter setting requirements and thereby eliminates the need of expensive preprocessing. 

Place, publisher, year, edition, pages
USA: Institute of Electrical and Electronics Engineers (IEEE), 2018
Keywords
Bayesian-nonparametrics, Dirichlet-process, short-text, text-analysis, topic-modeling, Information analysis, Bayesian nonparametrics, Dirichlet process, Short texts, Text analysis, Topic Modeling, Mixtures
National Category
Computer Sciences
Research subject
Skövde Artificial Intelligence Lab (SAIL)
Identifiers
urn:nbn:se:his:diva-16747 (URN)10.1109/ISCBI.2018.00025 (DOI)000462379700015 ()2-s2.0-85063024705 (Scopus ID)978-1-5386-9450-3 (ISBN)978-1-5386-9451-0 (ISBN)
Conference
6th International Symposium on Computational and Business Intelligence (ISCBI), 27-29 August 2018, Basel, Switzerland
Available from: 2019-04-05 Created: 2019-04-05 Last updated: 2019-09-30Bibliographically approved
Helldin, T., Steinhauer, H. J., Karlsson, A. & Mathiason, G. (2018). Situation Awareness in Telecommunication Networks Using Topic Modeling. In: 2018 21st International Conference on Information Fusion, FUSION 2018: . Paper presented at FUSION 2018 21st International Conference on Information Fusion, 10-13 July 2018, Cambridge, United Kingdom (pp. 549-556). IEEE
Open this publication in new window or tab >>Situation Awareness in Telecommunication Networks Using Topic Modeling
2018 (English)In: 2018 21st International Conference on Information Fusion, FUSION 2018, IEEE, 2018, p. 549-556Conference paper, Published paper (Refereed)
Abstract [en]

For an operator of wireless telecommunication networks to make timely interventions in the network before minor faults escalate into issues that can lead to substandard system performance, good situation awareness is of high importance. Due to the increasing complexity of such networks, as well as the explosion of traffic load, it has become necessary to aid human operators to reach a good level of situation awareness through the use of exploratory data analysis and information fusion techniques. However, to understand the results of such techniques is often cognitively challenging and time consuming. In this paper, we present how telecommunication operators can be aided in their data analysis and sense-making processes through the usage and visualization of topic modeling results. We present how topic modeling can be used to extract knowledge from base station counter readings and make design suggestions for how to visualize the analysis results to a telecommunication operator.

Place, publisher, year, edition, pages
IEEE, 2018
Keywords
Data handling, Data visualization, Information fusion, Design suggestions, Exploratory data analysis, Human operator, Information fusion techniques, Situation awareness, Telecommunication operators, Topic Modeling, Wireless telecommunications, Data mining
National Category
Computer and Information Sciences Computer Sciences Human Computer Interaction
Research subject
Skövde Artificial Intelligence Lab (SAIL); INF301 Data Science
Identifiers
urn:nbn:se:his:diva-16297 (URN)10.23919/ICIF.2018.8455529 (DOI)2-s2.0-85054066646 (Scopus ID)978-0-9964527-6-2 (ISBN)978-0-9964527-7-9 (ISBN)978-1-5386-4330-3 (ISBN)
Conference
FUSION 2018 21st International Conference on Information Fusion, 10-13 July 2018, Cambridge, United Kingdom
Available from: 2018-10-15 Created: 2018-10-15 Last updated: 2019-02-08Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0001-7106-0025

Search in DiVA

Show all publications