Group 3 - Forecasting.pdf

NUFV FORECASTING PROCESS, ACCURACY, & APPROACHES HM222 - GROUP 3  FORECASTING Topic Overview Forecasts are important in operations management because they predict future d...

NUFV FORECASTING PROCESS, ACCURACY, & APPROACHES HM222 - GROUP 3  FORECASTING Topic Overview Forecasts are important in operations management because they predict future demand. This helps ensure that supply matches demand, which is the main goal of operations management. STEPS IN THE FORECASTING PROCESS FORECAST ACCURACY APPROACHES TO FORECASTING GROUP 3  Steps in the Forecasting Process CHIGANE There are six Determine the Purpose of the Forecast basic steps in How will it be used and when will it be needed? This step will provide an indication of the level of detail required in the forecast, the amount of resources (personnel, computer time, the forecasting dollars) that can be justified, and the level of accuracy necessary. process: Establish a Time Horizon The forecast must indicate a time interval, keeping in mind that accuracy decreases as the time horizon increases. Obtain, Clean, and Analyze Appropriate Data Obtaining the data can involve significant effort. Once obtained, the data may need to be “cleaned” to get rid of outliers and obviously incorrect data before analysis.  Select a forecasting technique There are six Choose an appropriate method for forecasting based on the basic steps in data and the situation. Different techniques work better for different types of data and forecasting goals, such as qualitative methods for new products or quantitative the forecasting methods for well-established patterns. Make the forecast process: Use the chosen technique to create the forecast. This step involves applying the data and method to predict future outcomes. The forecast should align with the purpose and time horizon previously established. Monitor the forecast errors Check forecast errors to see if the forecast is accurate. If not, adjust the method, assumptions, or data, and create a new forecast. If demand is lower than expected, consider a price cut or promotion. If demand is higher, increase output through overtime, outsourcing, or other actions. Forecast Accuracy Understanding the Importance of Accurate Forecasting in Business Operations BIHASA  Why Forecast Accuracy Matters? Significance Forecast accuracy is essential for informed of Forecast decision-making in businesses. Accuracy Guides actions in inventory management, resource allocation, and strategic planning. While perfect accuracy is impossible due to complex variables, minimizing errors is key. Knowing potential forecast deviations helps in planning and maintaining operational efficiency.  Impact of Inaccurate Forecasts on Business Operations Overproduction / Misallocation of Customer Operational Underproduction Resources Dissatisfaction Challenges Leads to Causes Results from Creates excess inefficiencies delays or bottlenecks inventory or and higher wrong product and disrupts stockouts. costs. availability. workflows. Understanding Forecast Error What is Forecast Error? The difference between actual outcomes and forecasted values. Two Types of Forecast Error: Positive Error Forecast was too low (e.g., forecast = 90, actual = 100, error = +10). Negative Error Forecast was too high (e.g., forecast = 100, actual = 90, error = -10). Importance: Choosing Forecasting Techniques Assessing Forecast Performance  Summarizing Forecast Accuracy Forecast accuracy is a significant factor when deciding among forecasting alternatives. Accuracy is based on the historical error performance of a forecast. Mean Absolute Deviation (MAD): Average absolute error between actual and forecasted values. Simplest measure, treats all errors equally. Example: Let’s say you predicted sales for four months as 100, 150, 200, and 250 units, but the actual sales were 110, 140, 210, and 240 units. Errors: |110 - 100| = 10, |140 - 150| = 10, |210 - 200| = 10, |240 - 250| = 10 MAD = (10 + 10 + 10 + 10) / 4 = 10 MAD tells you that, on average, your forecast is off by 10 units. Mean Squared Error (MSE): Definition: Average of the squared errors, giving more weight to larger errors. Use: Ideal when larger errors are particularly impactful. Example: Using the same data as above (Forecast: 100, 150, 200, 250, 300; Actual: 110, 140, 210, 230, 320): Squared Errors: (110−100)²=100, (140−150)²=100, (210−200)²=100, (230−250)²=400, (320−300)²=400 MSE = (100 + 100 + 100 + 400 + 400) / 5-1 = 275 The MSE is 275. This means, on average, the squared errors between the forecasted and actual values are 275 units. Larger errors are given more weight because of the squaring process. Mean Absolute Percent Error (MAPE): Definition: Error expressed as a percentage of actual values. Use: Useful for comparing errors across different scales.  Choosing the Right Measure for Forecast Accuracy MAD MSE MAPE Ideal when all errors Suitable for comparing Best when large errors have similar errors across different are more detrimental. consequences. magnitudes. CONCLUSION: Operations managers should choose the measure that best aligns with organizational priorities. Example: Use MSE if large errors are costly, or MAPE for percentage-based comparisons. DISCUSSION BREAK ! ! ! ACTIVITY TIME ! Approaches to Forecasting CARALOS, BROSAS, BIHASA Qualitative Forecasting Qualitative forecasting methods rely on subjective inputs, which may include opinions, judgments, and intuition. This approach is particularly useful when there is little to no historical data available or when the situation is unprecedented. Key Features: Subjective: Based on opinions, judgments, and intuition. Flexible: Incorporates soft information that is difficult to quantify. Common Techniques: Executive Opinions: Collaboration among top managers for long-range planning. Salesforce Opinions: Insights from sales teams based on customer interactions, though potentially biased. Consumer Surveys: Direct feedback from consumers to gauge demand. Delphi Method: Consensus-building through iterative feedback from a panel of experts. Judgmental Forecasting Involve making predictions based on subjective inputs rather than strictly on numerical data. These inputs come from sources like consumer surveys, feedback from sales staff, opinions from managers and executives, and insights from panels of experts. These sources often provide valuable insights that are not captured by quantitative methods, offering a more nuanced understanding of potential future trends. Key Concepts: Based on Expert Judgment: Involves human inputs rather than strict data. Executive Opinions: Useful for long-term planning and new product development. Salesforce Opinions: Sales staff insights can indicate customer intentions. Consumer Surveys: Direct feedback from consumers can uncover valuable trends. Delphi Method: Achieves consensus through anonymous expert input over several rounds. Quantitative Forecasting Quantitative forecasting methods use mathematical models and historical data to predict future events. These methods are generally more objective and rely on hard data, making them less susceptible to personal biases Key Features: Objective Analysis: Relies on numerical data, reducing biases. Historical Data Usage: Projects future trends based on past data. Common Techniques: Time-Series Analysis: Predicts future events by analyzing past patterns (trends, seasonality, cycles). Associative Models: Uses relationships between variables to forecast outcomes (e.g., demand linked to price). Time-Series Forecasting Time-series forecasting analyzes a sequence of data points collected at consistent intervals. This approach assumes that the underlying patterns in the data will continue in the future. Key Concepts: Trend: Long-term direction of data. Seasonality: Regular short-term patterns (e.g., daily, weekly). Cycles: Long-term, wave-like variations tied to economic or political factors. Irregular Variations: Unusual events disrupting normal patterns (e.g., natural disasters). Random Variations: Residual fluctuations after accounting for other patterns. Time-Series Forecasting Common Techniques: Naive Methods: Uses the most recent data point for forecasting. Moving Averages: Smooths data by averaging recent points. Weighted Moving Averages: Assigns more weight to recent data. Exponential Smoothing: Adjusts forecasts based on previous errors with a smoothing constant. Associative Forecasting Techniques Also known as causal forecasting methods, are used to predict future values based on the relationship between the forecasted variable and one or more independent variables. These techniques assume that the variable you want to forecast is influenced by certain factors, and by understanding these relationships, you can make more accurate predictions. Key Concepts: Using Causal Variables: Forecasts based on relationships between variables. Example: Predicting paint demand based on price, advertising, and product features. Mathematical Models: Utilizes regression analysis and other statistical methods. Advanced Forecasting Methods Focus Forecasting: Combines multiple forecasting methods, selecting the most accurate one dynamically. Diffusion Models: Predicts adoption rates for new products based on the spread of similar products. Monitoring Forecast Errors Model Inadequacy: Missing variables or outdated models can cause errors. Irregular Variations: Unexpected events that disrupt normal patterns. Random Variations: The natural randomness in any data set.  CONCLUSION: Different forecasting approaches are suited for different situations. Combining qualitative and quantitative methods can enhance forecasting accuracy. Regularly monitoring and adjusting forecasts ensures they remain reliable. Accurate forecasts lead to better decision-making and business success.  NUFV THANK YOU FOR LISTENING! - GROUP 3  Group 3 "Great things in business are never done by one person; they're done by a team of people." — Steve Jobs Bihasa Chigane Shekinah Hope Shion Frias Brosas Caralos Harlee Ivette Maria Bea