Edinburgh Napier University

  This paper presents a thorough time series forecasting model intended to project future performance with respect to Sustainable Development Goal 1 (SDG 1) and the corresponding poverty index scores for the years 2024–2030, with a particular emphasis on the United States, Saudi Arabia, China, Egypt, and Sweden. A one-dimensional Convolutional Neural Network (1D-CNN) is used in the model to examine and extract patterns from historical socio-economic data from 2000 to 2022. The algorithm uses deep learning techniques to efficiently extract temporal correlations from the data, allowing for accurate predictions of each country's progress towards ending poverty and raising living standards. Because it can effectively handle time series data and find connections and patterns in earlier observations to guide future advancements in poverty alleviation tactics, the 1D-CNN architecture was chosen. The model was trained and validated using historical data to ensure predictions for the following years were based on real dynamics. To reflect the viability of reaching SDG 1 targets, forecasts were also limited to a realistic range of 0 to 100. The findings show that the model can correctly forecast shifts in poverty levels, which is consistent with expected worldwide patterns. These projections offer insightful information for international organisations, stakeholders, and policymakers involved in sustainability programs. By giving decision-makers the insight they need to deploy resources and carry out successful interventions effectively, the strategy speeds up the process of reaching SDG 1.