Alessandra Huydts
Thumbnail Image
Last Name
Huydts
First name
Alessandra
Email
alessandra.huydts@unisg.ch
Phone
+41 71 224 7405

Filters

6
Reset filters

Settings
Now showing 1 - 6 of 6
  • Publication
    Type: journal article
    Journal: Controlling : Zeitschrift für erfolgsorientierte Unternehmenssteuerung
    Volume: 34
    Issue: 2
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/109357
  • Publication
    Navigating Uncertainty: Human vs. Algorithmic Forecasting Approaches in Times of Crisis
    ( 2024-03-01)
    Alessandra Huydts
    ;
    Klaus Möller
    Forecasting plays a pivotal role in effective operational management, providing critical insights for decisiForecasting plays a pivotal role in effective operational management, providing critical insights for decision-makers. This study investigates the impact a crisis has on short-term forecasting performance, comparing human and algorithmic approaches. Drawing on data from a Swiss automotive distributor, the research distinguishes between four crisis types, where we focus on external sudden and external smoldering crises. Employing a mixed-method sequential explanatory research design, we find that algorithmic forecasts outperform human forecasts. Furthermore, variations in forecasting accuracy are observed across the different crisis situations. We find that independent of the specific business situation, overall, algorithmic forecasts outperform human forecasts. This study contributes valuable insights into the effectiveness of different forecasting methods in diverse crises, enhancing decision-making knowledge and resilience in times of uncertainty.on-makers. This study investigates the impact a crisis has on short-term forecasting performance, comparing human and algorithmic approaches. Drawing on data from a Swiss automotive distributor, the research distinguishes between four crisis types, where we focus on external sudden and external smoldering crises. Employing a mixed-method sequential explanatory research design, we find that algorithmic forecasts outperform human forecasts. Furthermore, variations in forecasting accuracy are observed across the different crisis situations. We find that independent of the specific business situation, overall, algorithmic forecasts outperform human forecasts. This study contributes valuable insights into the effectiveness of different forecasting methods in diverse crises, enhancing decision-making knowledge and resilience in times of uncertainty.
    Type: conference paper
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/119593
  • Publication
    From Data to Decision: A Comparative Study of Sales Forecasting Methods
    ( 2024-07-08)
    Alessandra Huydts
    With the advent of big data and advanced analytics, management accounting practices are increasingly interested in integrating sophisticated forecasting models to enhance decision-making processes. This study embarks on a comparative analysis to assess the appropriateness and performance of three modeling methodologies—traditional time series models, machine learning driver-based models, and the Facebook Prophet algorithm—in the context of sales forecasting for management accounting purposes. My research utilizes a mixed-methods approach, combining detailed quantitative and qualitative analyses of sales forecasting outcomes from multiple retail case studies. The findings reveal that the Facebook Prophet algorithm excels in direct (B2C) sales scenarios characterized by high data granularity and frequent data collection, offering superior accuracy and flexibility in adjusting for seasonal variations and market trends. Conversely, driver-based models demonstrate robust performance in indirect (B2B) sales scenarios where the available data does not directly reflect actual sales outcomes, such as situations with significant time lags or when data is aggregated excessively due to bulk purchases or sales. Traditional time series models serve as a useful baseline but generally underperform compared to the other methods. This study contributes to the management accounting literature by delineating the conditions under which each forecasting model has the highest efficacy, considering factors such as data availability, external influences, robustness to change, and data aggregation levels. The implications of this research extend to the design of forecasting systems, suggesting a nuanced approach to model selection that incorporates both technological capabilities and organizational dynamics.
    Type: working paper
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/120764
  • Publication
    Beyond Translation: Enabling Diverse Team Collaboration in Management Accounting Related Data Science Projects
    ( 2024-07-08)
    Alessandra Huydts
    ;
    Iuliana Sandu
    In the evolving field of management accounting, the integration of machine learning techniques is reshaping traditional forecasting methods, which have historically relied solely on human expertise. This shift not only necessitates a new set of skills but also transforms the work dynamic from individual effort to a focus on team-based communication and collaboration. This raises the question: How can effective communication and collaboration occur in teams with professional diversity, in effect, in teams where members ‘speak different languages’? This paper presents a case study from a Swiss automotive distributor's project aimed at advancing analytical liquidity forecasting through visual process mapping. This study explores how the meta-process of the visual process mapping (i.e., the process of the process) supports data science-based forecasting. By analyzing meeting transcripts and materials collected during the data-science-based liquidity forecasting project, we draw the conclusion that for management accounting-related data science projects, the process of creating visual process maps to support the data science task is instrumental in opening channels of collaboration by filling up the communication gaps presented in a diverse team.
    Type: working paper
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/120763
  • Publication
    Navigating Uncertainty: Human vs. Algorithmic Forecasting Approaches in Times of Crisis
    ( 2024-07-08)
    Alessandra Huydts
    ;
    Klaus Möller
    Forecasting plays a pivotal role in effective operational management, providing critical insights for decision-makers. This paper endeavors to discern the comparative performance of human and algorithmic forecasting, especially within crises, to test the resilience and adaptability of these methodologies. Drawing on data from a Swiss automotive distributor, the research distinguishes between four crisis types, where we focus on external sudden and external smoldering crises. Employing a mixed-method research design, we find that independent of the specific crisis situation, overall, algorithmic forecasts outperform human forecasts. However, for both human and algorithmic forecasts, variations in forecasting accuracy are observed, with smoldering and sudden crises exhibiting less accurate forecasts than non-crisis situations. This study contributes valuable insights into the effectiveness of different forecasting methods in diverse crises, enhancing decision-making knowledge and resilience in times of uncertainty. Furthermore, we hope that by showing the superiority of algorithmic forecasts and delineating their applicability, we can relieve decision-makers of the inherent distrust in algorithms.
    Type: working paper
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/120766
  • Publication
    Applying CRISP-DM in the Finance Function: A Cash Forecasting Case Study
    ( 2024-07-08)
    Alessandra Huydts
    ;
    Klaus Möller
    ;
    Michael Weiser
    Business analytics promises faster and more accurate cash forecasting results, but implementation is not trivial as the underlying business drivers need to be modeled. Thus, in this article, we elaborate on the established CRISP-DM process model for the implementation of business analytics use cases and apply it to the management accounting and finance context. We show how CRISP-DM can be applied to build a machine learning-based cash forecasting model based on a real-world case study at a large European car retailer. This will help managers improve and structure their business analytics initiatives.
    Type: working paper
    URI: https://www.alexandria.unisg.ch/handle/20.500.14171/120765