Abstract
Regulators of electricity distribution networks have typically applied Data Envelopment Analysis (DEA) to cross-section data for benchmarking purposes. However, the use of panel data to analyse the impact of regulatory policies on productivity change over time is less frequent. The main purpose of this paper is to construct a Malmquist productivity index to examine the recent productivity change experienced by Norwegian distribution companies between 2004 and 2007. The Malmquist index is decomposed in order to explore the sources of productivity change, and to identify the innovator companies that pushed the frontier forward each year. The input and output variables considered are those used by the Norwegian regulator. In order to reflect appropriately the exogenous conditions where the companies operate, the efficiency model used in this paper incorporates geography variables as outputs of the DEA model. Unlike the model used by the regulator, we included virtual weight restrictions in the DEA formulation to correct the biases in the DEA results that may be associated to a judicious choice of weights by some of the companies. Journal of the Operational Research Society (2012) 63, 982-990. doi: 10.1057/jors.2011.82 Published online 26 October 2011

Abstract
A good relationship between companies and customers is a crucial factor of competitiveness. Market segmentation is a key issue for companies to develop and maintain loyal relationships with customers as well as to promote the increase of company sales. This paper proposes a method for market segmentation in retailing based on customers' lifestyle, supported by information extracted from a large transactional database. A set of typical shopping baskets are mined from the database, using a variable clustering algorithm, and these are used to infer customers lifestyle. Customers are assigned to a lifestyle segment based on their purchases history. This study is done in collaboration with an European retailing company.

Abstract
Automatic Root Cause Analysis solutions aid analysts in finding problems' root causes by using automatic data analysis. When trying to locate the root cause of a problem in a manufacturing process, an issue-denominated overlap can occur. Overlap can impede automated diagnosis using algorithms, as the data make it impossible to discern the influence of each machine on the quality of products. This paper proposes a new measure of overlap based on an information theory concept called Positive Mutual Information. This new measure allows for a more detailed analysis. A new approach is developed for automatically finding the root causes of problems when overlap occurs. A visualization that depicts overlapped locations is also proposed to ease practitioners' analysis. The proposed solution is validated in simulated and real case-study data. Compared to previous solutions, the proposed approach improves the capacity to pinpoint a problem's root causes.

Abstract
PurposeThe paper aims to define a dashboard of indicators to assess the quality performance of higher education institutions (HEI). The instrument is termed SMART-QUAL.Design/methodology/approachTwo sources were used in order to explore potential indicators. In the first step, information disclosed in official websites or institutional documentation of 36 selected HEIs was analyzed. This first step also included in depth structured high managers' interviews. A total of 223 indicators emerged. In a second step, recent specialized literature was revised searching for indicators, capturing additional 302 indicators.FindingsEach one of the 525 total indicators was classified according to some attributes and distributed into 94 intermediate groups. These groups feed a debugging, prioritization and selection process, which ended up in the SMART-QUAL instrument: a set of 56 key performance indicators, which are grouped in 15 standards, and, in turn, classified into the 3 HEI missions. A basic model and an extended model are also proposed.Originality/valueThe paper provides a useful measure of quality performance of HEIs, showing a holistic view to monitor HEI quality from three fundamental missions. This instrument might assist HEI managers for both assessing and benchmarking purposes. The paper ends with recommendations for university managers and public administration authorities.

Abstract
Root cause analysis (RCA) is the process through which we find the true cause of a problem. It is a crucial process in manufacturing, as only after finding the root cause and addressing it, it is possible to improve the manufacturing operation. However, this is a very time-consuming process, especially if the amount of data about the manufacturing operation is considerable. With the increase in automation and the advent of Industry 4.0, sensorization of manufacturing environments has expanded, increasing with it the data available. The conjuncture described gives rise to the challenge and the opportunity of automatizing root cause analysis (at least partially), making this process more efficient, using tools from data mining and machine learning to help the analyst find the root cause of a problem. This paper presents an overview of the literature that has been published in the last 17 years on developing automatic root cause analysis (ARCA) solutions in manufacturing. The literature on the topic is disperse and it is currently lacking a connecting thread. As such, this study analyzes how previous studies developed the different elements of an ARCA solution for manufacturing: the types of data used, the methodologies, and the evaluation measures of the methods proposed. The proposed conceptualization establishes the base on which future studies on ARCA can develop results from this analysis, identifying gaps in the literature and future research opportunities.

Abstract
This paper describes a study on applying data mining techniques to power transformer failure prediction. The data set used consisted not only on DGA tests, but also in other tests done to the transformer’s insulating oil. This dataset presented several challenges, such as highly imbalanced classes (common in failure prediction problems), and the temporal nature of the observations. To overcome these challenges, several techniques were applied for prediction and better understand the dataset. Pre-processing and temporality incorporation in the dataset is discussed. For prediction, a 1-class and 2-class SVM, decision trees and random forests, as well as a LSTM neural network were applied to the dataset. As the prediction performance was low (high false-positive rate), we conducted a test to ascertain if the amount of data collected was sufficient. Results indicate that the frequency of data collection was not adequate, hinting that the degradation period was shorter than the periodicity of data collection.