Prediction of Weekly Price Trend of Garlic Based on Classification Algorithm and Combined Features

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The addresses manuscript important and current issues related to predicting the weekly garlic price trend based on the classification algorithm and combined features.

The literature review is incomplete. Please refer to more literature sources. Please include recent publications on similar topics.

The high proportion of similarities in the manuscript is disturbing (Percentage of match: 27%). Please pay attention to this, especially longer fragments of the text where similarities are indicated, e.g.: "The binomial logistic regression model is a classification model represented by a conditional probability distribution P(Y|X) in the form of a parameterized logistic distribution. Here, the random variable X takes the value of a real number and the random variable Y takes the value of 1 or 0."; "Here, x∊Rn is the input, Y∊{0,1} is the output, w∊Rn and b∊R are the parameters, w is called the weight vector, b is called the bias, and 𝒘 ∙ 𝒙 is the inner product of w and x. For a given input instance x, P(Y=1|x) and P(Y=0|x) can be obtained according to the formulas. Logistic regression compares the sizes of the two conditional probability values and assigns the instance x to the class with the larger probability value."; "Xgboost is a machine learning algorithm based on gradient boosting trees, which can be used for regression and classification problems. It improves the original GBDT model, making the model performance greatly enhanced. As a forward additive model, its core is to use the ensemble idea - Boosting idea, to integrate multiple weak learners into a strong learner by certain methods."

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1. Objective and Significance: The study focuses on predicting the weekly price trends of garlic, particularly in Jinxiang, China, to support the sustainable development of the garlic industry. The paper identifies the significant fluctuations in garlic prices and their impact on stakeholders, highlighting the need for accurate prediction methods to manage price risks effectively​​.

2. Methodology: The authors propose a new feature combination, termed De_Vo, derived from volatility data and sequences decomposed using the Variational Mode Decomposition (VMD) method. The study employs classification algorithms—logistic regression, SVM, and Xgboost—to predict price trends. The methodology also outlines how different features were extracted and combined to enhance prediction accuracy​​.

3. Results: The study's results show that the combined features improve prediction outcomes compared to using single features. In binary classification scenarios, the Xgboost model outperforms logistic regression and SVM in terms of accuracy. The paper demonstrates the effectiveness of the combined feature approach in capturing price trends and enhancing the predictive models' accuracy​​.

4. Conclusions and Implications: The paper concludes that the combined feature approach and the use of Xgboost offer significant improvements in predicting garlic price trends. The authors suggest that these methodologies could be applied to other agricultural products to assist in price fluctuation predictions and risk management. The study encourages further research into optimizing feature combinations and exploring additional algorithmic models to improve prediction accuracy​​.

5. Suggestions:

   • While the paper successfully demonstrates the applicability of machine learning techniques in agricultural economics, future research could extend to multi-regional studies to validate the model's generalizability.
   • The paper could benefit from a more in-depth discussion on the practical implications of the findings for farmers, traders, and policymakers, providing clearer guidelines on how to implement the proposed prediction models in real-world settings.

Comments on the Quality of English Language

1. Clarity and Conciseness: In several instances, sentences are overly lengthy and complex, which could hinder understanding. The paper would benefit from shorter, clearer sentences that directly convey the intended message without unnecessary complexity.

2. Grammar and Syntax: There are minor grammatical errors and awkward phrasings scattered throughout the document. These could be rectified through careful editing and proofreading to ensure that the paper meets the high standard expected in academic publishing.

3. Consistency: There are inconsistencies in terms, formatting, and style. Consistent terminology, especially when dealing with technical concepts, helps maintain clarity and professionalism. The paper should adhere to a specific style guide to improve its overall consistency.

4. Technical Jargon: While technical terms are expected in a scientific paper, it is important to ensure that they are clearly defined and used appropriately. Some sections of the paper might benefit from additional explanations or simplifications, particularly for readers who may not be experts in machine learning or agricultural economics.

5. Flow and Organization: The paper's overall flow can be improved to ensure that the narrative moves smoothly from one section to the next, guiding the reader through the research's background, methodology, results, and conclusions logically and intuitively.

6. Typographical Errors: The paper has minor typographical errors that, while not significantly detracting from its academic quality, do affect its professionalism. A thorough review and correction of these errors are recommended.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In the abstract appears “De”, “Vo” and “De_Vo” without any explanation. Please explain it.

Explain what are minor and major agricultural product market

Explain figure 1 in section 2.

Discussion is presented with only one reference. It is too poor.

In line 458, the accuracy of Xgboost reachs 72.9%. That means that it fails in 27.1%. If so, it is too high.

In lines 466 to 470, previous studies are indicated but they are not presented.

How can the author states that it can predict the fluctuation with so low level of accuracy (line 476)

 

Explain better lines 489-490 “feature does not fully take into account the price change factors.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The revised manuscript may be accepted in present form.