Athanasia Barka-Papadimitriou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Athanasia Barka-Papadimitriou has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Eleni Desiotou
Food Loss and Waste Prevention Unit, Municipality of Halandri, Solomou10 Str., Halandri, 15232 Athens, Greece
1 shared publication - Gerasimos Lyberatos
School of Chemical Engineering, National Technical University of Athens, 15780 Athens, Greece
1 shared publication - Kostas Efthimiou
Food Loss and Waste Prevention Unit, Municipality of Halandri, Solomou10 Str., Halandri, 15232 Athens, Greece
1 shared publication - Vassilis Lyberatos
School of Electrical and Computer Engineering, National Technical University of Athens, Iroon Polytechneiou 9, Zografou, 15780 Athens, Greece
1 shared publication
Latest Publications
- Open Access
A Machine Learning Approach for the Completion, Augmentation and Interpretation of a Survey on Household Food Waste Management
Households are the major contributor to food waste generation in the European Union according to the recently published data from Eurostat. Promoting food systems sustainability and aspiring to achieve the United Nations SDG 12.3 requires a better insight to the underlying drivers of the household food waste occurrence. The present study presents the combination of a well-established method of acquiring information, the questionnaire surveys, with a state-of-the-art technology for data imputation and interpretation using machine learning (ML). The Food Loss and Waste Prevention Unit (FLWPU) of the municipality of Halandri employed two surveys within the framework of the European funded projects Food Connections and FOODRUS. The first questionnaire was designed for rapid completion, to maximize response rates and minimize respondent burden, ensuring the collection of a consistent core dataset. A total of 154 replies were collected. The second questionnaire, associated with FOODRUS, was more detailed, enabling the participants to provide more in-depth information on their household food waste (HHFW) practices. In total, 43 responses were collected. ML algorithms were applied for data enhancement and data clustering. Specifically, ML and statistical techniques are applied for data imputations. An XGBoost algorithm was trained so as to capture complex relationships between variables. Behavioral intentions and effective strategies for reducing food waste at the community level are identified from the responses of both questionnaires, while a clustering of respondents in five groups emerged by using k-means, thus providing valuable insight into targeted HHFW prevention action plans.
Open AccessA Machine Learning Approach for the Completion, Augmentation and Interpretation of a Survey on Household Food Waste Management
Households are the major contributor to food waste generation in the European Union according to the recently published data from Eurostat. Promoting food systems sustainability and aspiring to achieve the United Nations SDG 12.3 requires a better insight to the underlying drivers of the household food waste occurrence. The present study presents the combination of a well-established method of acquiring information, the questionnaire surveys, with a state-of-the-art technology for data imputation and interpretation using machine learning (ML). The Food Loss and Waste Prevention Unit (FLWPU) of the municipality of Halandri employed two surveys within the framework of the European funded projects Food Connections and FOODRUS. The first questionnaire was designed for rapid completion, to maximize response rates and minimize respondent burden, ensuring the collection of a consistent core dataset. A total of 154 replies were collected. The second questionnaire, associated with FOODRUS, was more detailed, enabling the participants to provide more in-depth information on their household food waste (HHFW) practices. In total, 43 responses were collected. ML algorithms were applied for data enhancement and data clustering. Specifically, ML and statistical techniques are applied for data imputations. An XGBoost algorithm was trained so as to capture complex relationships between variables. Behavioral intentions and effective strategies for reducing food waste at the community level are identified from the responses of both questionnaires, while a clustering of respondents in five groups emerged by using k-means, thus providing valuable insight into targeted HHFW prevention action plans.