Background and Objective: Happiness is a universal fundamental human goal. Since the emergence of Positive Psychology, a major focus in psychological research has been to study the role of certain factors in the prediction of happiness. The conventional methodologies are based on linear relationships, such as the commonly used Multivariate Linear Regression (MLR), which may suffer from the lack of representative capacity to the varied psychological features. Using Deep Neural Networks (DNN), we define a Happiness Degree Predictor (H-DP) based on the answers to five psychometric standardized questionnaires.
Methods: A Data-Structure driven architecture for DNNs (D-SDNN) is proposed for defining a HDP in which the network architecture enables the conceptual interpretation of psychological factors associated to happiness. Four different neural network configurations have been tested, varying the number of neurons and the presence or absence of bias in the hidden layers. Two metrics for evaluating the influence of conceptual dimensions have been defined and computed: one quantifies the influence weight of the conceptual dimension in absolute terms and the other one pinpoints the direction (positive or negative) of the influence.
Materials: A cross-sectional survey targeting non-institutionalized adult population residing in Spain was completed by 823 cases. The total of 111 elements of the survey are grouped by socio-demographic data and by five psychometric scales (Brief COPE Inventory, EPQR-A, GHQ-28, MOS-SSS and SDHS) measuring several psychological factors acting one as the outcome (SDHS) and the four others as predictors.
Results: Our D-SDNN approach provided a better outcome (MSE: 1.46 · 10−2) than MLR (MSE: 2.30 · 10−2), hence improving by 37% the predictive accuracy, and allowing to simulate the conceptual structure.
Conclusions: We observe a better performance of Deep Neural Networks (DNN) with respect to traditional methodologies. This demonstrates its capability to capture the conceptual structure for predicting happiness degree through psychological variables assessed by standardized questionnaires. It also permits to estimate the influence of each factor on the outcome without assuming a linear relationship.
The pursuit of happiness is a universal - both cultural and time wise - core driver of human behaviour. Since ancient times pivotal and referent philosophical figures, as for example Aristotle1 from West or Zhuangzi2 from East, devoted much of their work to the idea of happiness as an ultimate purpose of human existence. The major proof that this consciousness pursuit of happiness should be considered as a fundamental human goal is the resolution adopted by the United Nations General Assembly on June 28th, 2012 where March, 20th was proclaimed the International Day of Happiness:
Recognizing the relevance of happiness and well-being as universal goals and aspirations in the lives of human beings around the world and the importance of their recognition in public policy objectives.
Recognizing also the need for a more inclusive, equitable and balanced approach to economic growth that promotes sustainable development, poverty eradication, happiness and the well-being of all peoples .
Consistent with this resolution, the United Nations (UN) has created a civilian based movement for a happier world [2,3], and took the lead to well-being and happiness as a principal aim in the development and launch of the 17 Sustainable Development Goals of the 2030 Agenda for Sustainable Development [4,5].
1.1. Happiness-Degree Predictor
Since the emergence of Positive Psychology  as the scientific study of factors that lead humans – both at the individual and collective level– to thrive, the research community has consistently built up the evidence-based knowledge about the so-called happiness or subjective well-being [7–14].
Happiness and depression are terms employed in daily life to denote affective states and mood swings, which are reliably represented as falling at opposite ends of a bipolar valence continuum [15,16]. For illustrative purposes, a graphical representation of the emotional valence spectrum is displayed in Fig. 1.
As it can be seen, depression is allocated at the very end of the negative affect side whereas happiness is placed at the opposite one. This implies that happiness is not just the absence of negative mood and affective states, but also the presence of positive ones.
Regarding happiness predictors, existent research has found psychological factors such as stress coping strategies [17,18], perceived social support [19–22] or personality [23–26] to have a considerable weight in its emergence. Up to now, the traditional methodological approach employed for happiness degree prediction has been a Multivariate Linear Regression (MLR) .
Emerging paradigms, novel approaches, and tools such as deep learning are becoming increasingly influential in psychological research as in the case of emotion recognition [28–30], sentiment analysis and/or classification [31–33]. It is worth to mention that both topics were endorsed in recent special issues in the last years [34–36] demonstrating the significance of the study and enabling us to avoid one of the pressing constraints of MLR that is the assumption of a linear relationship between the predictors (psychological factors) and the outcome (happiness degree).
Recent studies in sentiment analysis enclosed inside the field of psychology show the tendency to monitor the state of the people through social network activity, image/video and sentence classification [32,37–39]. These researches show the use of convolutional deep learning approaches which present a better behaviour for feature extraction and selection. Our study aims to mimic –without assuming any linear relationship– the structure of a set of psychometric scales which are conformed by structured data with prediction and interpretation purposes, becoming unnecessary the use of the convolutional technology because of the nature of data.