Ordinal Classification Problems Research Articles

Ultra Fast Deep Lane Detection With Hybrid Anchor Driven Ordinal Classification.

Modern methods mainly regard lane detection as a problem of pixel-wise segmentation, which is struggling to address the problems of efficiency and challenging scenarios like severe occlusions and extreme lighting conditions. Inspired by human perception, the recognition of lanes under severe occlusions and extreme lighting conditions is mainly based on contextual and global information. Motivated by this observation, we propose a novel, simple, yet effective formulation aiming at ultra fast speed and the problem of challenging scenarios. Specifically, we treat the process of lane detection as an anchor-driven ordinal classification problem using global features. First, we represent lanes with sparse coordinates on a series of hybrid (row and column) anchors. With the help of the anchor-driven representation, we then reformulate the lane detection task as an ordinal classification problem to get the coordinates of lanes. Our method could significantly reduce the computational cost with the anchor-driven representation. Using the large receptive field property of the ordinal classification formulation, we could also handle challenging scenarios. Extensive experiments on four lane detection datasets show that our method could achieve state-of-the-art performance in terms of both speed and accuracy. A lightweight version could even achieve 300+ frames per second(FPS). Our code is at https://github.com/cfzd/Ultra-Fast-Lane-Detection-v2.

ORFEO: Ordinal classifier and Regressor Fusion for Estimating an Ordinal categorical target

In this paper we present a novel methodology, referenced as ORFEO (Ordinal classifier and Regressor Fusion for Estimating an Ordinal categorical target), to enhance the performance in ordinal classification problems for which the latent variable is observable. ORFEO is an artificial neural network model incorporating two outputs, one for ordinal classification, using the cumulative link model, and one for regression, using a linear model. Both outputs are simultaneously optimised considering a loss function that linearly combines both classification and regression losses. The main motivation behind developing the proposed approach is to enhance the performance of a standard ordinal classifier. This improvement is facilitated by considering the regression output, which allows the model to differentiate between patterns within the same category. The ORFEO model is applied to two problems in the field of marine and ocean engineering: short-term prediction of both significant wave height and flux of energy. Both problems are addressed considering four different coastal zones of the United States of America, using 13 datasets formed by buoys measurements and reanalysis data. A comprehensive comparison against 20 methodologies, including regression and nominal/ordinal classification approaches is performed, by using diverse nominal and ordinal performance metrics. Ranks achieved indicate that ORFEO outperforms all the compared methodologies in terms of all the performance measures, demonstrating the efficacy and robustness of the proposal. Finally, a statistical analysis is conducted, concluding that there are statistically significant differences across ordinal and nominal performance metrics in favour of the proposed ORFEO model.

A general explicable forecasting framework for weather events based on ordinal classification and inductive rules combined with fuzzy logic

This paper presents a method for providing explainability in the integration of artificial intelligence (AI) and data mining techniques when dealing with meteorological prediction. Explainable artificial intelligence (XAI) refers to the transparency of AI systems in providing explanations for their predictions and decision-making processes, and contribute to improve prediction accuracy and enhance trust in AI systems. The focus of this paper relies on the interpretability challenges in ordinal classification problems within weather forecasting. Ordinal classification involves predicting weather phenomena with ordered classes, such as temperature ranges, wind speed, precipitation levels, and others. To address this challenge, a novel and general explicable forecasting framework, that combines inductive rules and fuzzy logic, is proposed in this work. Inductive rules, derived from historical weather data, provide a logical and interpretable basis for forecasting; while fuzzy logic handles the uncertainty and imprecision in the weather data. The system predicts a set of probabilities that the incoming sample belongs to each considered class. Moreover, it allows the expert decision-making process to be strengthened by relying on the transparency and physical explainability of the model, and not only on the output of a black-box algorithm. The proposed framework is evaluated using two real-world weather databases related to wind speed and low-visibility events due to fog. The results are compared to both ML classifiers and specific methods for ordinal classification problems, achieving very competitive results in terms of ordinal performance metrics while offering a higher level of explainability and transparency compared to existing approaches.

Ordinal classification for interval-valued data and interval-valued functional data

The aim of ordinal classification is to predict the ordered labels of the output from a set of observed inputs. Interval-valued data refers to data in the form of intervals. For the first time, interval-valued data and interval-valued functional data are considered as inputs in an ordinal classification problem. Six ordinal classifiers for interval data and interval-valued functional data are proposed. Three of them are parametric, one of them is based on ordinal binary decompositions and the other two are based on ordered logistic regression. The other three methods are based on the use of distances between interval data and kernels on interval data. One of the methods uses the weighted k-nearest-neighbor technique for ordinal classification. Another method considers kernel principal component analysis plus an ordinal classifier. And the sixth method, which is the method that performs best, uses a kernel-induced ordinal random forest. They are compared with naïve approaches in an extensive experimental study with synthetic and original real data sets, about human global development, and weather data. The results show that considering ordering and interval-valued information improves the accuracy. The source code and data sets are available at https://github.com/aleixalcacer/OCFIVD.

TactONet: Tactile Ordinal Network Based on Unimodal Probability for Object Hardness Classification

Hardness is one of the most critical tactile properties for robots to recognize objects. Machine learning methods have shown superior performance in object hardness classification. However, existing machine learning methods for tactile hardness classification cannot use the ordinal information between hardness classes because the one-hot encoding only cares about the correct class and ignores the inter-class relationship. To solve this problem, we propose to generalize the one-hot encoding using unimodal distributions including the Poisson and binomial distributions for tactile ordinal classification problems, resulting in two tactile ordinal networks (TacONet): TacONet-p and TacONet-b. Furthermore, we collect a tactile hardness dataset on the silicone samples with three different shapes (Shapes A, B, C), and each shape samples have thirteen hardness classes ranging from 0A (Shore A scale) to 60A at 5A intervals. We validate the resulting method for tactile hardness classification using a real robot. Experimental results demonstrate that compared with state-of-the-art methods, the proposed method achieves better classification performance in terms of accuracy and quadratic weighted kappa (QWK) on the tactile hardness dataset, reaching a classification accuracy up to 99.5% and a QWK up to 99.9% on Shape C. <i>Note to Practitioners</i>&#x2014;In the field of robotics tactile recognition, hardness classification is one of the most important and common tasks for robots to accurately recognize objects, particularly when the environment is dark or visual sensors are not working. In this paper, we propose a novel tactile ordinal network for tactile hardness classification tasks. The existing machine learning models for tactile hardness classification are trained by minimizing the cross-entropy loss between predicted vectors and one-hot encoding vectors of true classes, which makes the models only care about the correct classes and ignores the inter-class relationship of hardness classes. In other words, these models have the same probability to misclassify a hardness class with any other hardness class. To tackle this problem, we propose to generalize the one-hot encoding method using a unimodal distribution method to encode the true classes. The unimodal distribution encoding vectors can make the model learn the ordinal information between classes. It is proved that the proposed method is able to effectively improve the classification accuracy and QWK in a tactile hardness classification task.

#5803 ARTERIOVENOUS FISTULA BLOOD FLOW ESTIMATION THROUGH AI MODEL EXPLOITING THE PATIENT'S ROUTINELY COLLECTED DATA

Abstract Background and Aims Clinical monitoring and surveillance are key pillars of Arteriovenous Fistula (AVF) management in hemodialysis patients. They are aimed at limiting the risk of suboptimal dialysis dose and Vascular Access (VA) failure. AVF blood flow (Qa) is commonly used to assess VA function. Several methods have been proposed for the measurement of Qa including doppler ultrasound, and Body Thermal Monitor, among others. All these diagnostic tests are, in general, time-consuming for clinical staff, operator dependent, and costly for healthcare providers. To overcome limitations in the uptake of current Qa measurement techniques, we used data automatically recorded by dialysis machine sensors and medical information captured in electronic charts records to estimate Qa using a machine learning technique. Method For this historical cohort study, we analyzed electronic health records (EHR) of adult patients from four different European countries (Czech Republic, Portugal, Slovakia, and Spain), receiving in-center hemodialysis therapy in Fresenius Medical Care dialysis clinics between January 1st, 2015 and June 30th, 2022 registered in the European Clinical Database (EuCliD®). All patients consented their pseudo-anonymized data be used for secondary data analysis. The input dataset included 49 variables representing the health status of the patients, functional parameters of AVF function and the HD treatment parameters. The input variables were collected in the 90 days before the Qa measure. We constructed metrics representing the 90-day average and trend of each functional and medical parameter. Qa was classified in three levels: &amp;lt; 525 ml/h (very low), 525 ml/h - 925 ml/h (low), &amp;gt; 925 ml/h (normal). We estimated Qa as ordinal classification problem. Therefore, we used 2 binary classifiers based on the XGBoost algorithm. The estimation of the first-class, P(Qa&amp;lt;525), is given by the first classifier and the estimation of the third class, P(Qa≥925), is given by the second classifier. The probability of the middle class is computed as P(Qa≥525)⋅(1-P(Qa≥925)). The Qa estimation accuracy was assessed computing mean absolute error (MAE), precision and recall, F1-score and confusion matrix. Results Our dataset included 46,292 Qa measurements referred to 5,940 different hemodialysis patients. We obtained an overall precision of 0.77, a recall of 0.76, a F1-score of 0.76, a MAE of 0.27. The same metrics for each class are shown in Table 1. The model was able to detect fistula with “very low flow” with a precision of 0.9 and 16% of the missed “very low flow” AVFs are predicted as “low flow” and 4.7% are predicted as normal flow. The Confusion Matrix for each class is shown in Figure 1. Conclusion In this study we showed that clinically relevant Qa classes can be accurately predicted by resorting to routinely collected clinical data extracted from a electronic health record without any additional effort from healthcare professionals, training or instrumentation. Qa assessment is an important parameter in the evaluation of AVF function. Our algorithm accurately discriminated patients with “very low flow” that may be referred to vascular surgeon evaluation. It might help the AVF surveillance process, without adding time-consuming procedures for clinical staff or costs for healthcare provider.

Extracting relevant predictive variables for COVID-19 severity prognosis: An exhaustive comparison of feature selection techniques.

With the COVID-19 pandemic having caused unprecedented numbers of infections and deaths, large research efforts have been undertaken to increase our understanding of the disease and the factors which determine diverse clinical evolutions. Here we focused on a fully data-driven exploration regarding which factors (clinical or otherwise) were most informative for SARS-CoV-2 pneumonia severity prediction via machine learning (ML). In particular, feature selection techniques (FS), designed to reduce the dimensionality of data, allowed us to characterize which of our variables were the most useful for ML prognosis. We conducted a multi-centre clinical study, enrolling n = 1548 patients hospitalized due to SARS-CoV-2 pneumonia: where 792, 238, and 598 patients experienced low, medium and high-severity evolutions, respectively. Up to 106 patient-specific clinical variables were collected at admission, although 14 of them had to be discarded for containing ⩾60% missing values. Alongside 7 socioeconomic attributes and 32 exposures to air pollution (chronic and acute), these became d = 148 features after variable encoding. We addressed this ordinal classification problem both as a ML classification and regression task. Two imputation techniques for missing data were explored, along with a total of 166 unique FS algorithm configurations: 46 filters, 100 wrappers and 20 embeddeds. Of these, 21 setups achieved satisfactory bootstrap stability (⩾0.70) with reasonable computation times: 16 filters, 2 wrappers, and 3 embeddeds. The subsets of features selected by each technique showed modest Jaccard similarities across them. However, they consistently pointed out the importance of certain explanatory variables. Namely: patient's C-reactive protein (CRP), pneumonia severity index (PSI), respiratory rate (RR) and oxygen levels -saturation Sp O2, quotients Sp O2/RR and arterial Sat O2/Fi O2-, the neutrophil-to-lymphocyte ratio (NLR) -to certain extent, also neutrophil and lymphocyte counts separately-, lactate dehydrogenase (LDH), and procalcitonin (PCT) levels in blood. A remarkable agreement has been found a posteriori between our strategy and independent clinical research works investigating risk factors for COVID-19 severity. Hence, these findings stress the suitability of this type of fully data-driven approaches for knowledge extraction, as a complementary to clinical perspectives.

Soft labelling based on triangular distributions for ordinal classification

Recently, solving ordinal classification problems using machine learning and deep learning techniques has acquired important attention. There are many real-world problems in different areas of knowledge where a categorical variable needs to be predicted, and the existing categories follow an order associated with the nature of the problem: e.g. medical diagnosis with different states of a disease, or industrial quality assessment with different levels of quality. In these problems, it is quite common that the final label for each sample is determined by a group of experts with different opinions, and all opinions are usually summarised in a single crisp label by means of a given statistic (e.g. the median or the mode). Applying standard ordinal classifiers to these crisp labels could result in overfitting, as the labelling information is considered as totally certain. In this work, we propose a unimodal regularisation approach based on soft labelling, i.e. the ordinal information is used to introduce the inherent uncertainty of the label fusion. Specifically, said regularisation is based on using triangular distributions to simulate the aforementioned fusion of the expert opinions, where a parameter is used to decide the amount of probability that is assigned to the target category and the adjacent ones (according to the ordinal scale). The strategy could be applied to the loss function used by any ordinal classification learning algorithm, but we focus on deep learning in this paper. The proposal is compared to a baseline approach for nominal classification tasks and other state-of-the-art unimodal regularisation methods, and the experimental validation includes six benchmark datasets and five performance metrics. The results along with the statistical analysis show that the proposed methodology significantly outperforms the rest of the methods.

Deep learning based hierarchical classifier for weapon stock aesthetic quality control assessment

In the last years, multiple quality control tasks consist in classifying some items based on their aesthetic characteristics (aesthetic quality control, AQC), where usually the aspect of the material is not measurable and is based on expert observation. Given the increasing amount of images in this domain, deep learning (DL) models can be used to extract and classify the most discriminative patterns. Frequently, when trying to evaluate the quality of a manufactured product, the categories are naturally ordered, resulting in an ordinal classification problem. However, the ordinal categories assigned by an expert can be arranged in different levels that somehow model a hierarchy of the AQC task. In this work, we propose a DL approach to improve the classification performance in problems where categories are naturally ordered and follow a hierarchical structure. The proposed approach is evaluated on a real-world dataset that defines an AQC task and compared with other state-of-the-art DL methods. The experimental results show that our hierarchical approach outperforms the state-of-the-art ones.

Automating the decision making process of Todd’s age estimation method from the pubic symphysis with explainable machine learning

Age estimation is a fundamental task in forensic anthropology for both the living and the dead. The procedure consists of analyzing properties such as appearance, ossification patterns, and morphology in different skeletonized remains. The pubic symphysis is extensively used to assess adults’ age-at-death due to its reliability. Nevertheless, most methods currently used for skeleton-based age estimation are carried out manually, even though their automation has the potential to lead to a considerable improvement in terms of economic resources, effectiveness, and execution time. In particular, explainable machine learning emerges as a promising means of addressing this challenge by engaging forensic experts to refine and audit the extracted knowledge and discover unknown patterns hidden in the complex and uncertain available data. In this contribution we address the automation of the decision making process of Todd’s pioneering age assessment method to assist the forensic practitioner in its application. To do so, we make use of the pubic bone data base available at the Physical Anthropology lab of the University of Granada. The machine learning task is significantly complex as it becomes an imbalanced ordinal classification problem with a small sample size and a high dimension. We tackle it with the combination of an ordinal classification method and oversampling techniques through an extensive experimental setup. Two forensic anthropologists refine and validate the derived rule base according to their own expertise and the knowledge available in the area. The resulting automatic system, finally composed of 34 interpretable rules, outperforms the state-of-the-art accuracy. In addition, and more importantly, it allows the forensic experts to uncover novel and interesting insights about how Todd’s method works, in particular, and the guidelines to estimate age-at-death from pubic symphysis characteristics, generally.

SMAA-Choquet-FlowSort: A novel user-preference-driven Choquet classifier applied to supplier evaluation

The Choquet integral has been used as an aggregation operator to deal with interacting criteria in different types of problems. For ordinal classification problems, the majority of Choquet integral-based models proposed in the literature are built upon a supervised machine learning perspective, where a training data set is considered. Despite the effectiveness of these classification algorithms, their training-data-dependency may be considered a drawback in some decision-making problems. Our goal is then to propose a new multiple criteria Choquet classifier to conduct sorting based on user preference information. The classifier is built inside the FlowSort framework and, as such, aggregates intensity of preferences with respect to pairs of criteria instead of directly aggregating the criteria evaluations. This characteristic allows criteria and limiting profiles to be assessed by heterogeneous scales. In addition, we apply the Stochastic Multi-criteria Acceptability Analysis to the proposed classifier in order to elicit the Choquet capacities, model uncertain input data and analyze robustness of the results. The proposed classifier is applied to a real decision problem regarding the evaluation of pharmaceutical suppliers.

Siamese network based fine grained classification for Diabetic Retinopathy grading

Diabetic Retinopathy (DR) is a complication of diabetes which affects the retina. Early and the correct identification of the DR is necessary to determine the appropriate treatment. Automated DR grading is a fine-grained classification problem as some of the lesions are very small in size and are difficult to distinguish from non-DR regions. In this work, we propose a Siamese network based Convolutional Neural Network architecture for DR grading, which aims to improve the single eye based DR grading performance by incorporating the single eye based features with the patient-level DR features (i.e., features extracted from both eyes of the patient). To capture the patient-level DR information we propose two approaches: one is based on Bilinear pooling which aims to capture higher order statistical information by considering the correlation between the eyes of a patient, and the other one is based on simply averaging the features from both eyes. We experimentally show that both of the proposed approaches perform better than other approaches proposed in the literature for DR grading. In addition, as DR grading is an ordinal classification problem, we investigated the effect of different loss functions including the widely used Cross Entropy loss, Quadratic Weighted Kappa loss, Mean Squared Error Loss, and Ordinal Regression based loss, and show that Mean Squared Error loss gives the better system–annotator agreement. On the challenging, large scale, public Kaggle EyePACS dataset (consists of 88,702 images) our proposed approach achieves a Kappa score of 0.86 and an Accuracy value of 84.6%, indicating that the proposed approach is the new state-of-the-art.

A decision tree-based method for ordinal classification problems

In ordinal classification problems, the class value exhibits a natural order. Usually, these problems are solved as multiclass classification problems while discarding the ordering form of the class. Recently, several research studies have proposed novel methods for ordinal classification problems while aiming to predict a predefined objective value. These methods are based on modification of the splitting criteria of decision tree-based algorithms. These methods consider the ordinal nature of the data and the magnitude of the potential classification error from a predefined predicted objective. This research aims to consolidate these methods and generalize them for any decision tree-based method while constructing decision trees according to any predefined objective value calculated from the training dataset. Furthermore, to evaluate the proposed method, a comprehensive experimental study is performed based on known ordinal datasets. The proposed method is found to significantly outperform its counterpart nonordinal models. This observation confirms that ordering information benefits ordinal methods and improves their performance. Additionally, the ordinal decision tree-based methods achieved competitive performance compared to state-of-the-art ordinal techniques. The results are validated with various performance measures that are commonly used for ordinal and non-ordinal classification problems.

Unimodal regularisation based on beta distribution for deep ordinal regression

Currently, the use of deep learning for solving ordinal classification problems, where categories follow a natural order, has not received much attention. In this paper, we propose an unimodal regularisation based on the beta distribution applied to the cross-entropy loss. This regularisation encourages the distribution of the labels to be a soft unimodal distribution, more appropriate for ordinal problems. Given that the beta distribution has two parameters that must be adjusted, a method to automatically determine them is proposed. The regularised loss function is used to train a deep neural network model with an ordinal scheme in the output layer. The results obtained are statistically analysed and show that the combination of these methods increases the performance in ordinal problems. Moreover, the proposed beta distribution performs better than other distributions proposed in previous works, achieving also a reduced computational cost.

Aggregating user preferences in group recommender systems: A crowdsourcing approach

We present that group recommendations are similar to crowdsourcing, where the responses of different crowd workers are aggregated in the absence of ground truth. With this in mind, we mimic the use of the EM algorithm as in crowdsourcing to aggregate the preferences of group members to estimate group ratings and the expertise levels the group members. Moreover, for the first time in the literature, we cast the problem of estimating group rating as an ordinal classification problem relying on the natural ordering between the ratings, which allows us to define the expertise levels of the members in terms of sensitivity and specificity. In fact, we impose priors on the sensitivity and the specificity scores corresponding to the members, taking a Bayesian approach. We validate the effectiveness of the proposed aggregation method using the CAMRa2011 dataset, which consists of small and established groups, and the MovieLens dataset, which consists of large and random groups.

A comparative study of machine learning methods for ordinal classification with absolute and relative information

The performance of an ordinal classifier is highly affected by the amount of absolute information (labelled data) available for training. In order to make up for a lack of sufficient absolute information, an effective way out is to consider additional types of information. In this work, we focus on ordinal classification problems that are provided with additional relative information. We augment several classical machine learning methods by considering both absolute and relative information as constraints in the corresponding optimization problems. We compare these augmented methods on popular benchmark datasets. The experimental results show the effectivenesses of these methods for combining absolute and relative information.

Cumulative link models for deep ordinal classification

This paper proposes a deep convolutional neural network model for ordinal regression by considering a family of probabilistic ordinal link functions in the output layer. The link functions are those used for cumulative link models, which are traditional statistical linear models based on projecting each pattern into a 1-dimensional space. A set of ordered thresholds splits this space into the different classes of the problem. In our case, the projections are estimated by a non-linear deep neural network. To further improve the results, we combine these ordinal models with a loss function that takes into account the distance between the categories, based on the weighted Kappa index. Three different link functions are studied in the experimental study, and the results are contrasted with a statistical analysis. The experiments run over two different ordinal classification problems and the statistical tests confirm that these models improve the results of a nominal model and outperform other robust proposals considered in the literature.

Joint maximization of accuracy and information for learning the structure of a Bayesian network classifier

Although recent studies have shown that a Bayesian network classifier (BNC) that maximizes the classification accuracy (i.e., minimizes the 0/1 loss function) is a powerful tool in both knowledge representation and classification, this classifier: (1) focuses on the majority class and, therefore, misclassifies minority classes; (2) is usually uninformative about the distribution of misclassifications; and (3) is insensitive to error severity (making no distinction between misclassification types). In this study, we propose to learn the structure of a BNC using an information measure (IM) that jointly maximizes the classification accuracy and information, motivate this measure theoretically, and evaluate it compared with six common measures using various datasets. Using synthesized confusion matrices, twenty-three artificial datasets, seventeen UCI datasets, and different performance measures, we show that an IM-based BNC is superior to BNCs learned using the other measures—especially for ordinal classification (for which accounting for the error severity is important) and/or imbalanced problems (which are most real-life classification problems)—and that it does not fall behind state-of-the-art classifiers with respect to accuracy and amount of information provided. To further demonstrate its ability, we tested the IM-based BNC in predicting the severity of motorcycle accidents of young drivers and the disease state of ALS patients—two class-imbalance ordinal classification problems—and show that the IM-based BNC is accurate also for the minority classes (fatal accidents and severe patients) and not only for the majority class (mild accidents and mild patients) as are other classifiers, providing more informative and practical classification results. Based on the many experiments we report on here, we expect these advantages to exist for other problems in which both accuracy and information should be maximized, the data is imbalanced, and/or the problem is ordinal, whether the classifier is a BNC or not. Our code, datasets, and results are publicly available http://www.ee.bgu.ac.il/~boaz/software .

Automatic Assessment of Fabric Smoothness Appearance Based on a Compact Convolutional Neural Network With Label Smoothing

In the textile and apparel industry, it remains a challenging task to evaluate the fabric smoothness objectively. In existing studies, the objective fabric smoothness assessment is defined as a typical image classification problem. However, the fabric smoothness labels contain sequence information, and the problem shall be defined as an ordinal classification problem. This article presents an effective method including an image preprocessing algorithm, a compact convolutional neural network(CNN) model, and a label smoothing process. Compared with the commonly used CNN frameworks, the proposed compact CNN model is more suitable for this small-sample and low-abstraction problem. The image processing algorithm can improve the model’s illumination adaptability, and the label smoothing process can modify the model to satisfy the ordinal classification problems better. In the experiments, the method is tested on a fabric image set including 385 graded fabric specimens. Within a 10-fold cross validation, the proposed method achieves 84.00%, 95.38%, and 100% average accuracies under errors of 0 degree, 0.5 degree, and 1 degree respectively. Implementation discussions on preprocessing and label smoothing verify their effectiveness in improving model performance in assessment accuracies and illumination stability. The proposed method outperforms the state-of-the-art methods for fabric smoothness assessment and a series of widely used deep learning methods. Promisingly, the proposed method can provide novel research ideas for the image-based fabric smoothness assessment.

Chronological Age Estimation Under the Guidance of Age-Related Facial Attributes

Although the researches of facial attributes’ analysis have been launched for decades, the estimation of chronological age attribute remains a big challenge. Previous researchers have found that some facial attributes (e.g., gender and race attributes) have close connections with the age attribute and make age estimation under a specific condition decided by various combinations of those age-related attributes which should be more reasonable. In this paper, we propose a generic framework based on a convolutional neural network, which can consider different conditions for age estimation and jointly output age and age-related facial attributes in the end. Compared with conventional methods, it is more efficient and universal. Besides, we view age estimation as a special multi-class ordinal classification problem and use a losses combination function to optimize the predicted probability distribution of individual age classes. These operations further improve the performance of age estimation. Finally, the proposed method achieves state-of-the-art results on both controlled and wild face datasets.