arXiv subjects

Eric Malmi

Publications and source records attributed to Eric Malmi.

27 records · Page 2Linked to original sources

arXiv2018

Computationally Inferred Genealogical Networks Uncover Long-Term Trends in Assortative Mating

Genealogical networks, also known as family trees or population pedigrees, are commonly studied by genealogists wanting to know about their ancestry, but they also provide a valuable resource for disciplines such as digital demography, genetics, and computational social science. These networks are typically constructed by hand through a very time-consuming process, which requires comparing large numbers of historical records manually. We develop computational methods for automatically inferring large-scale genealogical networks. A comparison with human-constructed networks attests to the accuracy of the proposed methods. To demonstrate the applicability of the inferred large-scale genealogical networks, we present a longitudinal analysis on the mating patterns observed in a network. This analysis shows a consistent tendency of people choosing a spouse with a similar socioeconomic status, a phenomenon known as assortative mating. Interestingly, we do not observe this tendency to consistently decrease (nor increase) over our study period of 150 years.

cs.SI↗

arXiv2018

Automatic Prediction of Discourse Connectives

Accurate prediction of suitable discourse connectives (however, furthermore, etc.) is a key component of any system aimed at building coherent and fluent discourses from shorter sentences and passages. As an example, a dialog system might assemble a long and informative answer by sampling passages extracted from different documents retrieved from the Web. We formulate the task of discourse connective prediction and release a dataset of 2.9M sentence pairs separated by discourse connectives for this task. Then, we evaluate the hardness of the task for human raters, apply a recently proposed decomposable attention (DA) model to this task and observe that the automatic predictor has a higher F1 than human raters (32 vs. 30). Nevertheless, under specific conditions the raters still outperform the DA model, suggesting that there is headroom for future improvements.

cs.CL↗

arXiv2017

Active Network Alignment: A Matching-Based Approach

Network alignment is the problem of matching the nodes of two graphs, maximizing the similarity of the matched nodes and the edges between them. This problem is encountered in a wide array of applications-from biological networks to social networks to ontologies-where multiple networked data sources need to be integrated. Due to the difficulty of the task, an accurate alignment can rarely be found without human assistance. Thus, it is of great practical importance to develop network alignment algorithms that can optimally leverage experts who are able to provide the correct alignment for a small number of nodes. Yet, only a handful of existing works address this active network alignment setting. The majority of the existing active methods focus on absolute queries ("are nodes $a$ and $b$ the same or not?"), whereas we argue that it is generally easier for a human expert to answer relative queries ("which node in the set $\{b_1, \ldots, b_n\}$ is the most similar to node $a$?"). This paper introduces two novel relative-query strategies, TopMatchings and GibbsMatchings, which can be applied on top of any network alignment method that constructs and solves a bipartite matching problem. Our methods identify the most informative nodes to query by sampling the matchings of the bipartite graph associated to the network-alignment instance. We compare the proposed approaches to several commonly-used query strategies and perform experiments on both synthetic and real-world datasets. Our sampling-based strategies yield the highest overall performance, outperforming all the baseline methods by more than 15 percentage points in some cases. In terms of accuracy, TopMatchings and GibbsMatchings perform comparably. However, GibbsMatchings is significantly more scalable, but it also requires hyperparameter tuning for a temperature parameter.

cs.SI↗

arXiv2017

Domain Adaptation for Resume Classification Using Convolutional Neural Networks

We propose a novel method for classifying resume data of job applicants into 27 different job categories using convolutional neural networks. Since resume data is costly and hard to obtain due to its sensitive nature, we use domain adaptation. In particular, we train a classifier on a large number of freely available job description snippets and then use it to classify resume data. We empirically verify a reasonable classification performance of our approach despite having only a small amount of labeled resume data available.

cs.CV↗

arXiv2016

DopeLearning: A Computational Approach to Rap Lyrics Generation

Writing rap lyrics requires both creativity to construct a meaningful, interesting story and lyrical skills to produce complex rhyme patterns, which form the cornerstone of good flow. We present a rap lyrics generation method that captures both of these aspects. First, we develop a prediction model to identify the next line of existing lyrics from a set of candidate next lines. This model is based on two machine-learning techniques: the RankSVM algorithm and a deep neural network model with a novel structure. Results show that the prediction model can identify the true next line among 299 randomly selected lines with an accuracy of 17%, i.e., over 50 times more likely than by random. Second, we employ the prediction model to combine lines from existing songs, producing lyrics with rhyme and a meaning. An evaluation of the produced lyrics shows that in terms of quantitative rhyme density, the method outperforms the best human rappers by 21%. The rap lyrics generator has been deployed as an online tool called DeepBeat, and the performance of the tool has been assessed by analyzing its usage logs. This analysis shows that machine-learned rankings correlate with user preferences.

cs.LG↗

arXiv2016

You Are What Apps You Use: Demographic Prediction Based on User's Apps

Understanding the demographics of app users is crucial, for example, for app developers, who wish to target their advertisements more effectively. Our work addresses this need by studying the predictability of user demographics based on the list of a user's apps which is readily available to many app developers. We extend previous work on the problem on three frontiers: (1) We predict new demographics (age, race, and income) and analyze the most informative apps for four demographic attributes included in our analysis. The most predictable attribute is gender (82.3 % accuracy), whereas the hardest to predict is income (60.3 % accuracy). (2) We compare several dimensionality reduction methods for high-dimensional app data, finding out that an unsupervised method yields superior results compared to aggregating the apps at the app category level, but the best results are obtained simply by the raw list of apps. (3) We look into the effect of the training set size and the number of apps on the predictability and show that both of these factors have a large impact on the prediction accuracy. The predictability increases, or in other words, a user's privacy decreases, the more apps the user has used, but somewhat surprisingly, after 100 apps, the prediction accuracy starts to decrease.

cs.SI↗

arXiv2012

Semi-Supervised Anomaly Detection - Towards Model-Independent Searches of New Physics

Most classification algorithms used in high energy physics fall under the category of supervised machine learning. Such methods require a training set containing both signal and background events and are prone to classification errors should this training data be systematically inaccurate for example due to the assumed MC model. To complement such model-dependent searches, we propose an algorithm based on semi-supervised anomaly detection techniques, which does not require a MC training sample for the signal data. We first model the background using a multivariate Gaussian mixture model. We then search for deviations from this model by fitting to the observations a mixture of the background model and a number of additional Gaussians. This allows us to perform pattern recognition of any anomalous excess over the background. We show by a comparison to neural network classifiers that such an approach is a lot more robust against misspecification of the signal MC than supervised classification. In cases where there is an unexpected signal, a neural network might fail to correctly identify it, while anomaly detection does not suffer from such a limitation. On the other hand, when there are no systematic errors in the training data, both methods perform comparably.

physics.data-an↗

arXiv2010

Soft Classification of Diffractive Interactions at the LHC

Multivariate machine learning techniques provide an alternative to the rapidity gap method for event-by-event identification and classification of diffraction in hadron-hadron collisions. Traditionally, such methods assign each event exclusively to a single class producing classification errors in overlap regions of data space. As an alternative to this so called hard classification approach, we propose estimating posterior probabilities of each diffractive class and using these estimates to weigh event contributions to physical observables. It is shown with a Monte Carlo study that such a soft classification scheme is able to reproduce observables such as multiplicity distributions and relative event rates with a much higher accuracy than hard classification.

hep-ex↗

arXiv2009

Multivariate Techniques for Identifying Diffractive Interactions at the LHC

Close to one half of the LHC events are expected to be due to elastic or inelastic diffractive scattering. Still, predictions based on extrapolations of experimental data at lower energies differ by large factors in estimating the relative rate of diffractive event categories at the LHC energies. By identifying diffractive events, detailed studies on proton structure can be carried out. The combined forward physics objects: rapidity gaps, forward multiplicity and transverse energy flows can be used to efficiently classify proton-proton collisions. Data samples recorded by the forward detectors, with a simple extension, will allow first estimates of the single diffractive (SD), double diffractive (DD), central diffractive (CD), and non-diffractive (ND) cross sections. The approach, which uses the measurement of inelastic activity in forward and central detector systems, is complementary to the detection and measurement of leading beam-like protons. In this investigation, three different multivariate analysis approaches are assessed in classifying forward physics processes at the LHC. It is shown that with gene expression programming, neural networks and support vector machines, diffraction can be efficiently identified within a large sample of simulated proton-proton scattering events. The event characteristics are visualized by using the self-organizing map algorithm.

hep-ex↗