Towards a sustainable artificial intelligence: A case study of energy efficiency in decision tree algorithms

2021 ◽  
Author(s):  
Mariza Ferro ◽  
Gabrieli D. Silva ◽  
Felipe B. Paula ◽  
Vitor Vieira ◽  
Bruno Schulze
Keyword(s):  
Artificial Intelligence ◽  
Energy Efficiency ◽  
Decision Tree ◽  
Tree Algorithms

Creating an AI-Conversational Agent in Psychotherapy: Recognizing User Intents

2021 ◽  
Author(s):  
Olga Troitskaya ◽  
Andrey Zakharov
Keyword(s):  
Artificial Intelligence ◽  
Decision Tree ◽  
Interpersonal Relations ◽  
Emotional State ◽  
Conversational Agent ◽  
Intent Recognition ◽  
Self Help ◽  
User Query ◽  
Tree Algorithms ◽  
Commit Suicide

In recent years there has been a growth of psychological chatbots performing important functions from checking symptoms to providing psychoeducation and guiding self-help exercises. Technologically these chatbots are based on traditional decision-tree algorithms with limited keyword recognition. A key challenge to the development of conversational artificial intelligence is intent recognition or understanding the goal that the user wants to accomplish. The user query on psychological topic is often emotional, highly contextual and non goal-oriented, and therefore may contain vague, mixed or multiple intents. In this study we made an attempt to identify and categorize user intents with relation to psychological topics using the database of 43 000 messages from iCognito Anti-depression chatbot. We have identified 24 classes of user intents that can be grouped into larger categories, such as: a) intents to improve emotional state; b) intents to improve interpersonal relations; c) intents to improve physical condition; d) intents to solve practical problems; e) intents to make a decision; f) intents to harm oneself or commit suicide; g) intent to blame or criticize oneself. This classification may be used for the development of conversational artificial intelligence in the field of psychotherapy.

Penerapan Metode Klasifikasi Decision Tree dan Algoritma C4.5 dalam Memprediksi Kriteria Nasabah Kredit Mega Auto Finance

2020 ◽  
Vol 7 (2) ◽  
pp. 200
Author(s):  
Puji Santoso ◽  
Rudy Setiawan
Keyword(s):  
Data Mining ◽  
Decision Tree ◽  
Microsoft Excel ◽  
Customer Data ◽  
Data Mining Techniques ◽  
C4.5 Algorithm ◽  
Marketing Costs ◽  
Excel Format ◽  
Data Mining Application

One of the tasks in the field of marketing finance is to analyze customer data to find out which customers have the potential to do credit again. The method used to analyze customer data is by classifying all customers who have completed their credit installments into marketing targets, so this method causes high operational marketing costs. Therefore this research was conducted to help solve the above problems by designing a data mining application that serves to predict the criteria of credit customers with the potential to lend (credit) to Mega Auto Finance. The Mega Auto finance Fund Section located in Kotim Regency is a place chosen by researchers as a case study, assuming the Mega Auto finance Fund Section has experienced the same problems as described above. Data mining techniques that are applied to the application built is a classification while the classification method used is the Decision Tree (decision tree). While the algorithm used as a decision tree forming algorithm is the C4.5 Algorithm. The data processed in this study is the installment data of Mega Auto finance loan customers in July 2018 in Microsoft Excel format. The results of this study are an application that can facilitate the Mega Auto finance Funds Section in obtaining credit marketing targets in the future

Using the private finance initiative for energy efficiency projects at the urban scale

2016 ◽  
Vol 10 (1) ◽  
pp. 99-117 ◽  
Author(s):  
Alberto De Marco ◽  
Giulio Mangano ◽  
Fania Valeria Michelucci ◽  
Giovanni Zenezini
Keyword(s):  
Energy Efficiency ◽  
Urban Areas ◽  
Light Emitting Diode ◽  
Project Finance ◽  
Private Finance Initiative ◽  
Content Type ◽  
Light Emitting ◽  
Alternate Forms ◽  
Private Investors

Purpose – The purpose of this paper is to suggest the usage of the project finance (PF) scheme as a suitable mechanism to fund energy efficiency projects at the urban scale and present its advantages and adoption barriers. Design/methodology/approach – A case study is developed to renew the traffic lighting system of an Italian town via replacement of the old lamps with new light-emitting diode (LED) technology. Several partners are involved in the case project to construct a viable PF arrangement. Findings – The case study presents the viability of the proposed PF scheme that provides for acceptable financial returns and bankability. However, it also shows that the need for short concession periods may call for a public contribution to the initial funding to make the project more attractive to private investors. Practical implications – This case study is a useful guideline for governments and promoters to using the PF arrangement to fund energy efficiency investments in urban settings. It helps designing an appropriate PF scheme and understanding the advantages of PF to reduce risk and, consequently, increase the debt leverage and profitability of energy efficiency projects. Originality/value – This paper contributes to bridging the gap about the lack of works addressing the implementation of the PF mechanism in the energy efficiency sector in urban areas. The importance of this paper is also associated with the shortage of traditional public finance faced by many cities that forces to seek for alternate forms of financing.

Sign in / Sign up

Export Citation Format

Share Document