Subgroup discovery (SD) enables one to elicit patterns that strongly discriminate a class label. When it comes to numerical data, most of the existing SD approaches perform data discretizations and thus suffer from information loss. A few algorithms avoid such a loss by considering the search space of every interval pattern built on the dataset numerical values and provide an “anytime” property: at any moment, they are able to provide a result that improves over time. Given a sufficient time/memory budget, they may eventually complete an exhaustive search. However, such approaches are often intractable when dealing with high-dimensional numerical data, for instance, when extracting features from real-life multivariate time series. To overcome such limitations, we propose MonteCloPi, an approach based on a bottom-up exploration of numerical patterns with a Monte Carlo Tree Search. It enables to have a better exploration-exploitation trade-off between exploration and exploitation when sampling huge search spaces. Our extensive set of experiments proves the efficiency of MonteCloPi on high-dimensional data with hundreds of attributes. We finally discuss the actionability of discovered subgroups when looking for skill analysis from Rocket League action logs.

The need of predictive maintenance comes with an increasing number of incidents reported by monitoring systems and equipment/software users. In the front line, on-call engineers (OCEs) have to quickly assess the degree of severity of an incident and decide which service to contact for corrective actions. To automate these decisions, several predictive models have been proposed, but the most efficient models are opaque (say, black box), strongly limiting their adoption. In this paper, we propose an efficient black box model based on 170K incidents reported to our company over the last 7 years and emphasize on the need of automating triage when incidents are massively reported on thousands of servers running our product, an ERP. Recent developments in eXplainable Artificial Intelligence (XAI) help in providing global explanations to the model, but also, and most importantly, with local explanations for each model prediction/outcome. Sadly, providing a human with an explanation for each outcome is not conceivable when dealing with an important number of daily predictions. To address this problem, we propose an original data-mining method rooted in Subgroup Discovery, a pattern mining technique with the natural ability to group objects that share similar explanations of their black box predictions and provide a description for each group. We evaluate this approach and present our preliminary results which give us good hope towards an effective OCE's adoption. We believe that this approach provides a new way to address the problem of model agnostic outcome explanation.

Subgroup discovery (SD) is the task of discovering interpretable patterns in the data that stand out w.r.t. some property of interest. Discovering patterns that accurately discriminate a class from the others is one of the most common SD tasks. Standard approaches of the literature are based on local pattern discovery, which is known to provide an overwhelmingly large number of redundant patterns. To solve this issue, pattern set mining has been proposed: instead of evaluating the quality of patterns separately, one should consider the quality of a pattern set as a whole. The goal is to provide a small pattern set that is diverse and well-discriminant to~the target class. In this work, we introduce a novel formulation of the task of diverse subgroup set discovery where both discriminative power and diversity of the subgroup set are incorporated in the same quality measure. We propose an efficient and parameter-free algorithm dubbed FSSD and based on a greedy scheme. FSSD uses several optimization strategies that enable to efficiently provide a high quality pattern set in a short amount of time.

It is extremely useful to exploit labeled datasets not only to learn models but also to improve our understanding of a domain and its available targeted classes. The so-called subgroup discovery task has been considered for a long time. It concerns the discovery of patterns or descriptions, the set of supporting objects of which have interesting properties, e.g., they characterize or discriminate a given target class. Though many subgroup discovery algorithms have been proposed for transactional data, discovering subgroups within labeled sequential data and thus searching for descriptions as sequential patterns has been much less studied. In that context, exhaustive exploration strategies can not be used for real-life applications and we have to look for heuristic approaches. We propose the algorithm SeqScout to discover interesting subgroups (w.r.t. a chosen quality measure) from labeled sequences of itemsets. This is a new sampling algorithm that mines discriminant sequential patterns using a multi-armed bandit model. It is an anytime algorithm that, for a given budget, finds a collection of local optima in the search space of descriptions and thus subgroups. It requires a light configuration and it is independent from the quality measure used for pattern scoring. Furthermore, it is fairly simple to implement. We provide qualitative and quantitative experiments on several datasets to illustrate its added-value.

The success of electronic sports (eSports), where professional gamers participate in competitive leagues and tournaments, brings new challenges for the video game industry. Other than fun, games must be difficult and challenging for eSports professionals but still easy and enjoyable for amateurs. In this article, we consider Multi-player Online Battle Arena games (MOBA) and particularly, "Defense of the Ancients 2", commonly known simply as DOTA2. In this context, a challenge is to propose data analysis methods and metrics that help players to improve their skills. We design a data mining-based method that discovers strategic patterns from historical behavioral traces: Given a model encoding an expected way of playing (the norm), we are interested in patterns deviating from the norm that may explain a game outcome from which player can learn more efficient ways of playing. The method is formally introduced and shown to be adaptable to different scenarios. Finally, we provide an experimental evaluation over a dataset of 10 000 behavioral game traces.

Video game is a very lucrative industry, unleashed by the ubiquity of gaming devices, multi-player networks and live broadcasting platforms. Games generate large amounts of behavioural data which are valuable to face the new challenges of video game analytics such as detecting balance issues, bugs and cheaters. In electronic sports (e-sports), cyberathletes conceal their online training using different aliases or avatars (virtual identities), which allow them not being recognized by the opponents they may face in future competitions (with cash prices challenging already most of the traditional sports). It was recently suggested that behavioural data generated by the games allows predicting the avatar associated to a game play with high accuracy. However, when a player uses several avatars, accuracy drastically drops as prediction models cannot easily differentiate the player's different avatar aliases. Since mappings between players and avatars do not exist, we introduce the avatar aliases identification problem and propose an original approach for alias resolution based on supervised classification and Formal Concept Analysis. We thoroughly evaluate our method with the video game Starcraft 2 which has a very wide and active community with players from diverse cultures and nations. We show that under some circumstances, the avatars of a given player can easily be recognized as such. These results are valuable for e-sport structures (to help preparing tournaments), and game editors (detecting cheaters or usurpers).

To describe the dynamics taking place in networks that structurally change over time, we propose an approach to search for attributes whose value changes impact the topology of the graph. In several applications, it appears that the variations of a group of attributes are often followed by some structural changes in the graph that one may assume they generate. We formalize the triggering pattern discovery problem as a method jointly rooted in sequence mining and graph analysis. We apply our approach on three real-world dynamic graphs of different natures - a co-authoring network, an airline network, and a social bookmarking system - assessing the relevancy of the triggering pattern mining approach.

Recent developments in the frequent pattern mining framework uses additional measures of interest to reduce the set of discovered patterns. We introduce a rigorous and efficient approach to mine statistically significant, unexpected patterns in sequences of item sets. The proposed methodology is based on a null model for sequences and on a multiple testing procedure to extract patterns of interest. Experiments on sequences of replays of a video game demonstrate the scalability and the efficiency of the method to discover unexpected game strategies.

This paper shows how to embed a similarity relation between complex descriptions in concept lattices. We formalize similarity by a tolerance relation: objects are grouped within a same concept when having similar descriptions, extending the ability of FCA to deal with complex data. We propose two different approaches.~A first classical manner defines a discretization procedure. A second way consists in representing data by pattern structures, from which a pattern concept lattice can be constructed directly. In this case, considering a tolerance relation can be mathematically defined by a projection in a meet-semi-lattice. This allows to use concept lattices for their knowledge representation and reasoning abilities without transforming data. We show finally that resulting lattices are useful for solving information fusion problems.