Machine learning models have proven to be reliable methods in classification tasks. However, little research has been done on classifying dwelling characteristics based on smart meter & weather data before. Gaining insights into dwelling characteristics can be helpful to create/improve the policies for creating new dwellings at NZEB standard. This paper compares the different machine learning algorithms and the methods used to correctly implement the models. These methods include the data pre-processing, model validation and evaluation. Smart meter data was provided by Groene Mient, which was used to train several machine learning algorithms. The models that were generated by the algorithms were compared on their performance. The results showed that Recurrent Neural Network (RNN) 2performed the best with 96% of accuracy. Cross Validation was used to validate the models, where 80% of the data was used for training purposes and 20% was used for testing purposes. Evaluation metrices were used to produce classification reports, which can indicate which of the models work the best for this specific problem. The models were programmed in Python.
Machine learning models have proven to be reliable methods in classification tasks. However, little research has been conducted on the classification of dwelling characteristics based on smart meter and weather data before. Gaining insights into dwelling characteristics, which comprise of the type of heating system used, the number of inhabitants, and the number of solar panels installed, can be helpful in creating or improving the policies to create new dwellings at nearly zero-energy standard. This paper compares different supervised machine learning algorithms, namely Logistic Regression, Support Vector Machine, K-Nearest Neighbor, and Long-short term memory, and methods used to correctly implement these algorithms. These methods include data pre-processing, model validation, and evaluation. Smart meter data, which was used to train several machine learning algorithms, was provided by Groene Mient. The models that were generated by the algorithms were compared on their performance. The results showed that the Long-short term memory performed the best with 96% accuracy. Cross Validation was used to validate the models, where 80% of the data was used for training purposes and 20% was used for testing purposes. Evaluation metrics were used to produce classification reports, which indicates that the Long-short term memory outperforms the compared models on the evaluation metrics for this specific problem.
Elk jaar organiseert het Leerhuis van het Projectmanagementbureau samen met de Hogeschool van Amsterdam een thematische seminarreeks.De koppeling van de wetenschappelijke theorie met de dagelijkse praktijk biedt een waardevolle verdieping van de kennis van de medewerkers van het Projectmanagementbureau. Centraal staat de toepasbaarheid in ons werk aan complexe en multidisciplinaire opgaven in Amsterdam.In 2018/2019 hebben we gekozen voor het smart city-concept, waarbij big data en tools roepen om toepassingsmogelijkheden in de aanpak van stedelijke problematiek.Wie in de media de discussie over de opkomst van smart cities volgt, krijgt vaak het beeld voorgeschoteld van een grote controlekamer, vol met monitors en schermen. Daarop worden allerhande processen in de stad op de minuut gevolgd: de doorstroming van het verkeer op de hoofdwegen, de weersverwachting en waterstanden, meldingen aan politie en brandweer of samenscholingen van menigtes in de openbare ruimte. Naast deze ‘control room’-visie staan twee minder bekende benaderingen: ‘Smart Citizens’ zet technologie in om burgers meer zeggenschap te geven, terwijl de ‘Creative City’ de stad beziet als ‘living lab’. Alle drie de verschijningsvormen van de smart city bieden kansen om de kwaliteit van leven in de stad te verbeteren, én roepen tegelijkertijd vragen op over hun doelmatigheid en legitimiteit.De verschillende bijdragen aan de seminarreeks over smart cities gingen dieper in op de mogelijkheden en valkuilen van de drie smart city perspectieven. Deze zijn gebundeld in dit magazine en aangevuld met een aantal lessen en handvatten voor de medewerkers van het PMB.
MULTIFILE
The Dutch floriculture is globally leading, and its products, knowledge and skills are important export products. New challenges in the European research agenda include sustainable use of raw materials such as fertilizer, water and energy, and limiting the use of pesticides. Greenhouse growers however have little control over crop growth conditions in the greenhouse at individual plant level. The purpose of this project, ‘HiPerGreen’, is to provide greenhouse owners with new methods to monitor the crop growth conditions in their greenhouse at plant level, compare the measured growth conditions and the measured growth with expected conditions and expected growth, to point out areas with deviations, recommend counter-measures and ultimately to increase their crop yield. The main research question is: How can we gather, process and present greenhouse crop growth parameters over large scale greenhouses in an economical way and ultimately improve crop yield? To provide an answer to this question, a team of university researchers and companies will cooperate in this applied research project to cover several different fields of expertise The application target is floriculture: the production of ornamental pot plants and cut flowers. Participating companies are engaged in the cultivation of pot plans, flowers and suppliers of greenhouse technology. Most of the parties fall in the SME (MKB) category, in line with the RAAK MKB objectives.Finally, the Demokwekerij and Hortipoint (the publisher of the international newsletter on floriculture) are closely involved. The project will develop new knowledge for a smart and rugged data infrastructure for growth monitoring and growth modeling in the greenhouse. In total the project will involve approximately 12 (teacher) researchers from the universities and about 60 students, who will work in the form of internships and undergraduate studies of interesting questions directly from the participating companies.
The postdoc candidate, Sondos Saad, will strengthen connections between research groups Asset Management(AM), Data Science(DS) and Civil Engineering bachelor programme(CE) of HZ. The proposed research aims at deepening the knowledge about the complex multidisciplinary performance deterioration prediction of turbomachinery to optimize cleaning costs, decrease failure risk and promote the efficient use of water &energy resources. It targets the key challenges faced by industries, oil &gas refineries, utility companies in the adoption of circular maintenance. The study of AM is already part of CE curriculum, but the ambition of this postdoc is that also AM principles are applied and visible. Therefore, from the first year of the programme, the postdoc will develop an AM material science line and will facilitate applied research experiences for students, in collaboration with engineering companies, operation &maintenance contractors and governmental bodies. Consequently, a new generation of efficient sustainability sensitive civil engineers could be trained, as the labour market requires. The subject is broad and relevant for the future of our built environment being more sustainable with less CO2 footprint, with possible connections with other fields of study, such as Engineering, Economics &Chemistry. The project is also strongly contributing to the goals of the National Science Agenda(NWA), in themes of “Circulaire economie en grondstoffenefficiëntie”,”Meten en detecteren: altijd, alles en overall” &”Smart Industry”. The final products will be a framework for data-driven AM to determine and quantify key parameters of degradation in performance for predictive AM strategies, for the application as a diagnostic decision-support toolbox for optimizing cleaning &maintenance; a portfolio of applications &examples; and a new continuous learning line about AM within CE curriculum. The postdoc will be mentored and supervised by the Lector of AM research group and by the study programme coordinator(SPC). The personnel policy and job function series of HZ facilitates the development opportunity.
Publieke ambtenaren bij gemeenten zien in toenemende mate de relatie met de burger veranderen in de eigen werkomgeving. In een informatie-intensieve netwerksamenleving zien ambtenaren veel kansen in het toepassen van social media en smart-city-technologie. Open data, social media en high-tech media kunnen, indien strategisch toegepast, helpen om slimmer samen te gaan werken met burgers en bedrijven. Maar ambtenaren worstelen met het effectief toepassen van deze nieuwe digitale mogelijkheden voor participatie. De veranderingen in samenleving vragen bovendien om meer dan technologie alleen. Ze vragen om integrale benaderingen van digitale strategie inclusief vernieuwende werkvormen en aangepaste competenties. Vanuit de disciplines van smart city en elektronische participatie is er jarenlang onderzoek gedaan naar oplossingen voor het participatievraagstuk maar het is nog onduidelijk welke digitale strategieën het beste aansluiten bij de behoeften van deze tijd. Dit project beoogt een concreet werkinstrumentarium te ontwikkelen voor professionals van de lokale overheid om participatie te organiseren. Het gaat om het uitvinden van geschikte combinaties van technologie en communicatiemethoden voor effectieve participatie in de stad, dorp of buurt. Door het uittesten van diverse digitale strategieën in ‘Living-labs’ bij gemeenten willen we de know-how bundelen in een toolkit en beschikbaar maken voor professionals die werkzaam zijn in het publieke veld.
