<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Ryker on Medium]]></title>
        <description><![CDATA[Stories by Ryker on Medium]]></description>
        <link>https://medium.com/@ryker_65837?source=rss-1b573bb17e68------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*2OgBhwEgmNQ2vYSGj43YkA.jpeg</url>
            <title>Stories by Ryker on Medium</title>
            <link>https://medium.com/@ryker_65837?source=rss-1b573bb17e68------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 06:20:33 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ryker_65837/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Measuring the uncertainty of NFTs through Confidence Intervals.]]></title>
            <link>https://medium.com/nftbank-ai/measuring-the-uncertainty-of-nfts-through-confidence-intervals-dd089756fea?source=rss-1b573bb17e68------2</link>
            <guid isPermaLink="false">https://medium.com/p/dd089756fea</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[web3]]></category>
            <category><![CDATA[blockchain]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[nft]]></category>
            <dc:creator><![CDATA[Ryker]]></dc:creator>
            <pubDate>Wed, 07 Jun 2023 04:42:38 GMT</pubDate>
            <atom:updated>2024-04-04T15:58:02.190Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>TL’DR: Confidence intervals provide assistance in measuring the uncertainty of assets.</strong></h3><p>The estimation of volatility for NFT (Non-Fungible Token) assets is of significant interest in the domain of financial analysis. Volatility, representing the degree of price fluctuation and uncertainty, plays a crucial role in assessing the risk and potential return associated with these unique digital assets. <strong>In this context, confidence intervals can act as a valuable tool for measuring uncertainty and establishing a range of possible outcomes.</strong></p><p><strong>By computing confidence intervals for NFT asset volatility, we can unlock several opportunities for analysis and decision-making.</strong> Here, we explore those opportunities in the following ways:</p><ol><li><strong>Risk Management/ Portfolio Optimization</strong>: Confidence intervals enable investors to construct portfolios with a consideration of NFT asset volatility. By accounting for the uncertainty associated with each asset, investors can seek to optimize portfolio returns while managing risk. Confidence intervals facilitate the evaluation of trade-offs between risk and reward, supporting the identification of optimal asset allocation, position sizing, and hedging techniques</li><li><strong>Detailed Investment Decision-Making</strong>: By <em>comparing</em> confidence intervals across different NFT assets or investment opportunities, investors can make more informed investment decisions. Assets with narrower confidence intervals may indicate relatively lower volatility and potentially offer a more stable investment choice.</li><li><strong>Pricing and Valuation</strong>: Confidence intervals can aid in pricing NFT assets, especially in marketplaces where transaction data is limited or volatile. By incorporating confidence intervals into pricing models, market participants can better capture the potential price range within which an NFT asset’s value may fluctuate, allowing for more accurate valuation assessments.</li></ol><p>We aim to estimate confidence intervals for NFT asset volatility, providing a comprehensive framework for understanding and quantifying the uncertainty surrounding these unique digital assets.</p><h3>Why is Interval Estimation More Effective Than Point Estimation?</h3><p>NFTBank’s Valuation model aims to estimate the current asset value of NFTs using transaction data. Estimation can generally be divided into two types: point estimation and interval estimation. Point estimation involves estimating a single value, providing a concise and easily interpretable result.However, the limitation of point estimation lies in its inability to adequately reflect uncertainty based solely on the estimated value. To address this limitation, interval estimation is used to evaluate the reliability and uncertainty of the estimation more accurately.</p><p>Unlike point estimation, which aims to estimate a single point, interval estimation presents results as a range, acknowledging uncertainty and providing information about the possible values the estimated quantity can take with a certain level of confidence. This approach offers more statistical information by explicitly indicating the range of values that can be attributed to the estimated value.</p><p>Generally, interval estimation involves iterative estimation of estimation errors and establishing intervals based on assumptions about the distribution of the estimated value. However, such processes often require substantial computational resources, prompting us to explore alternative approaches for estimation. Therefore, based on these considerations, we intend to introduce quantile regression as a means to estimate intervals in our analysis.</p><h3>How Did We Enable Interval Estimation?</h3><p>To enable interval estimation, we employed the concept of quantile regression, which involves estimating quantiles of a distribution. Quantile regression provides a more robust estimation approach by directly estimating quantiles instead of relying solely on point estimates. This leads to more reliable and accurate interval estimations, considering the uncertainty and variability in the data.</p><p>It is known that the parameters and corresponding predictions that minimize the check loss represent the predictions for the tau-quantiles. Therefore, to obtain the desired confidence interval at a given significance level alpha, we constructed a model that simultaneously estimates the lower, medium, upper quantiles with respect to significance level alpha.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YTOX9MPZ8tdHtrz9X0Lx6Q.png" /><figcaption>An intuitive example for our confidence interval</figcaption></figure><p>However, many non-linear models, including neural networks, face a critical issue in estimating multiple quantiles simultaneously, known as the <em>crossing problem</em>.</p><p>The crossing problem refers to the occurrence of inversions between the predicted values of two quantiles. This means that the 40% quantile may have a higher value than the 60% quantile, which violates the properties of quantiles. When inversions occur, the range of values for the predicted interval cannot be preserved, necessitating pre- or post-hoc adjustments.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0UgS3fua2VwPXuqx7D08zQ.png" /><figcaption>An example for crossing problem. Some of those lower quantiles are cross over upper quantiles.</figcaption></figure><p>At NFTBank, we developed a model structure to address the crossing problem. Firstly, we created a custom objective function capable of estimating multiple quantiles, enabling us to utilize first-order optimization by computing the gradients. Additionally, we leveraged the monotonicity condition within the LightGBM module to ensure that the monotonicity condition holds for the estimated quantiles.</p><p>By implementing these methods, we successfully enabled interval estimation, allowing us to estimate quantiles and construct reliable confidence intervals for NFT asset valuation and volatility analysis.</p><h3>Conclusion</h3><p>For this article, we covered the importance of volatility in the NFT pricing ecosystem and how we manage the influence of volatility through our NFT Estimate Confidence Interval. Through a proper confidence interval, users can get a range of uncertainties compared to a single point to manage risk for different digital assets and optimize their portfolios.</p><p>The confidence interval is opened for free for a limited time only, before it goes to the Pro+ version. Test it out right now on our <a href="https://developer.nftbank.ai/reference/nft_confidence_interval_estimated_price_v1_nft__assetcontract___tokenid__interval_estimate_get">API Docs page</a>.</p><h3>Join our socials</h3><ul><li>Official website: <a href="https://nftbank.ai/">https://nftbank.ai</a></li><li>Twitter: <a href="https://twitter.com/NFTbank_ai">https://twitter.com/NFTbank_ai</a></li><li>Discord: <a href="https://t.co/N2kvmpvko6">discord.gg/8Rg63UnhYb</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dd089756fea" width="1" height="1" alt=""><hr><p><a href="https://medium.com/nftbank-ai/measuring-the-uncertainty-of-nfts-through-confidence-intervals-dd089756fea">Measuring the uncertainty of NFTs through Confidence Intervals.</a> was originally published in <a href="https://medium.com/nftbank-ai">NFTBank.ai</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to accurately estimate the price of individual NFTs — Part 2]]></title>
            <link>https://medium.com/nftbank-ai/how-to-accurately-estimate-the-price-of-individual-nfts-part-2-ac73e88aa3ad?source=rss-1b573bb17e68------2</link>
            <guid isPermaLink="false">https://medium.com/p/ac73e88aa3ad</guid>
            <dc:creator><![CDATA[Ryker]]></dc:creator>
            <pubDate>Wed, 13 Apr 2022 08:04:11 GMT</pubDate>
            <atom:updated>2024-04-04T16:24:41.212Z</atom:updated>
            <content:encoded><![CDATA[<h3>How to accurately estimate the price of individual NFTs — Part 2</h3><h3>Recap of Part I &amp; Preview of Part 2</h3><p>Many rely on an item’s latest transaction price or the Floor Price of the collection for NFT valuation. However, they do not accurately capture the value of an item because there is a significant variance within an NFT collection and high sparsity in the NFT trade dataset. (This is comparable to challenges faced when buying real estate, particularly those with few transaction records.)</p><p>Despite the challenge, this is a crucial problem to solve. Accurate NFT valuation — NFT price estimates close to the transaction price — gives valuable information to NFT collectors, whether they are managing their portfolio or assessing opportunities to buy or sell NFTs.</p><p><a href="https://medium.com/nftbank-ai/how-to-value-items-in-nft-projects-part-1-a87be8bcb21d">Our previous article</a> gave you a sneak peek at how NFTBank solves this problem. We introduced some of the difficulties in constructing a valuation model or the conventional time-serial model on NFT projects with low transaction volume, which leads to sparsity in trade datasets and extreme price change.</p><p>We also discussed the nuances in estimating NFT price in bundle sales, which occurs quite frequently, and how categorical traits value can be converted into meaningful numerical data, not just mapping one variable to one constant such as “Gold” mapping to 1.</p><p>We developed our baseline model that guarantees a highly accurate price estimation for hundreds of different NFT projects that address these difficulties and nuances. This model is what allowed us to serve numerous users and partners.</p><p>Since the model described in Part I, we have tested and applied various state-of-the-art ML models to improve model performance. In doing so, we have made significant strides in prediction accuracy. In this article, we share what the improvement looks like, the additional problem we needed to solve, and how we solved it to get to where we are now.</p><h3>What Effective NFT Price Estimate Looks Like</h3><p>Let’s address the most important question before we start: <em>how effective is the NFTBank’s price estimate</em>? Take a look at the predicted values and actual sales values of the BAYC collection from 2022–02–01 to 2022–04–06; the blue box plots indicate our price estimate values, whereas the red box plots indicate actual sales price. You can see our predicted prices are in close range with the actual sales price, even when certain traits such as ‘gold fur’ deviate far from the average.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Csio0DV7QKmrPGWf" /><figcaption>Boxplot of pred vs actual -BAYC</figcaption></figure><h3>How did we get here?</h3><h4>Low volume problem &amp; high price volatility continued to pose challenge</h4><p>The <strong>low volume problem</strong> as well as the <strong>high price volatility problem</strong> still lingered as challenge in applying ML-based prediction to NFT assets.</p><p><strong>Low volume problem</strong> led to over or underestimation of NFTs with rare traits. This is because few unordinary trades can skew the model significantly. This turned out to be the greatest challenge, which we describe further below.</p><p><strong>High price volatility</strong> led to incorrect estimation due to the nature of the ML model. ML model learns through past data and if a sudden movement in price occurs, it takes time for the model to absorb and learn, which results in lagging estimation.</p><p>Solving the low volume problem enabled us to better predict values of rarer NFTs which can drastically affect asset efficiency of super rare NFTs, while solving for high price volatility allowed our model to more closely track actual price movements. In this article, we’ll focus on the low volume problem, and address the high price volatility problem in our next one.</p><h3>Solving low volume problem in NFT ML prediction with Trait Reconstruction</h3><p>We’ve found that the key to handling the challenge of low transaction volume is tracking and processing price information of “<strong>similar</strong>” items. In other words, we need to follow the prices of comparable NFTs that are sold, determine how similar they are, and find appropriate prices. This is important because if similarity between items can be determined, the information from sparse data can be approximated with that of similar items.</p><p>In finding the similarity between NFTs, we worked with the most intuitive but core key principle that the most critical factor in determining the price of an NFT is the “<strong>trait</strong>”.</p><p>Here is an oversimplified version of what we do:</p><ol><li><strong>decompose the traits of NFT items</strong></li><li><strong>regroup NFT items by trait</strong></li><li><strong>compute the price statistics per each group</strong></li><li><strong>reconstruct the value of an NFT item using the computed statistics</strong></li></ol><p>It may look simple and pretty easy to implement, but enabling a model to recognize and learn the principle may not be as simple as you think. For example, just putting the trait in the model often causes the under-estimation problem since most of the transactions stay near the floor (which leads the ML model to estimate floor values; Because it is the efficient way to minimize the objective function).</p><p>In the NFT market, Steps 1–3 mentioned above can be easily solved by thinking of traits as properties. We can simply calculate the statistics for each trait by subdividing the trait possessed by the item. However, the challenge is in properly collecting the statistics for individual traits and turn them into the stat that can represent an item. In other words, we need to figure out <strong>which traits significantly impact price</strong> and then <strong>reconstruct trait information into items appropriately</strong>. Just because a few NFTs with ‘Orange’ background got sold at high price, it does not guarantee that all other NFTs with the ‘Orange’ background will also sell at high price.</p><p>Let’s take a look at a real life example. In BAYC, apes with “Solid Gold” fur have a high price range compared to apes with “Brown” fur. It implies that the process of decomposing the traits of NFT items (Step 1), regrouping NFT items by trait (Step 2), and compute the price statistics per each group (Step 3) can be solved directly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*SHMynFVO-eKPe1gI" /><figcaption>Time-serial price trend of apes with two traits; (Red) Solid gold, (Blue) Brown; x-axis: timestamp, y-axis: log(price)</figcaption></figure><p>However, the problem arises in (Step 4): how to reconstruct the value of an NFT item using the computed statistics? For example, consider apes with “Bored unshaven” and “Bored pizza” mouth that is possessed by 3690 apes and 30 apes, respectively. Looking at the price trend for NFTs with each trait, we see that most of the times, the price of NFTs with the “Bored pizza” are higher most of the times, but at times, NFTs with “Bored unshaven” are sold at higher price than the other.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*mXKQOSsUV7Zn3OCI" /><figcaption>Time-serial price trend of apes with Bored unshaven and Bored Pizza mouth: x-axis: timestamp, y-axis: log(price)</figcaption></figure><p>The points where the inversions occurred were when those NFTs had other valuable traits such as “Solid Gold”, “Trippy”, or “Black suit” apes. This example pinpoints the difficulty in addressing step 4, that is, effectively buildIng item information from a set of trait information.</p><h3>How did NFTBank solve it?</h3><h4>TL’DR</h4><p>We solved this challenge by devising a model that automatically detects key traits and brings together individual trait information to form item-specific information. We introduced the logic into the model and confirmed that the principle was helpful for the performance through numerous experiments. For example, the graph below shows the effectiveness of the trait reconstruction we’ve done. You can see that the predicted price (red line) closely follows the sold price trend, much more effectively than the project’s floor price trend.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*BO3Tu_lc3NnkFBgY" /><figcaption>Time-serial trend of item price prediction based on trait reconstruction; x-axis: timestamp, y-axis: log(price); (blue): sold price, (red): item price prediction based on trait reconstruction, (green): project floor price</figcaption></figure><h3>Full Version</h3><p>In general, in machine learning territory, response variable is often transformed with logarithmic, power transformation, or further, Box-Cox transformation. The transformations have a variety of purposes, but generally they want the data to satisfy some statistical assumptions. For example, in a pricing modeling, you can adjust the skewness of the target variable distribution by using logarithmic transformation.</p><p>We found that the price of NFT is not only skewed even if logarithmic transformation is taken, but also there are many cases where multiple modals occur. Thus, we developed a new method inspired by forward variable selection, and Box-Cox transformation to gather statistics from each trait.</p><p>The forward variable selection is one of the traditional statistical variable selection method. It begins with a model that contains no variables, in our case, traits. Then, it starts adding the most significant variables one after the other until a pre-specified stopping rule is reached. Here, we defined the stopping rule as the mean absolute error (MAE) that is used equally in the objective function of the model.</p><p>The approximated item price is constructed through a weighted average of information of the selected traits (in this case, the weight is applied differently depending on the time, and the more recent the data is, the larger). Therefore, the approximated item price calculated through our transformation behaves very similar to the actual selling price.</p><p>The strength of our ML system is able to correct the sparsity by grasping and considering all trends in each trait. In other words, our model can estimate the item with the transaction has not occurred at all, or even if the same combination of traits has never been sold.</p><h3>Limitation</h3><p>Our fundamental approach is to approximate the price of the item to the price of the trait. Naturally, it is impossible to estimate the trait that has never been traded (but, who knows? we may correct ourselves 😉 ). It also implies that our method is vulnerable to wash-trading scenarios. If the same NFT has been traded many times with incomprehensible (too-high or too-low) price, our trait information will be affected no matter how robust it is. We had experienced a temporary decline in model performance due to the wash-trading as we thought.</p><h3>Conclusion</h3><p>The model performance of NFTBank is constantly being improved through numerous experiments and efforts. All team members are working day and night to make NFTBank the most effective tool for managing NFT assets. A big thank to our users for supporting us and giving us great feedback. This is a huge motivator for us and means a lot to us.</p><p>Please reach out to us here if you want to use our Price Estimate for your product or service. We love to partner!</p><h3>Join our socials</h3><ul><li>Official website: <a href="https://nftbank.ai/">https://nftbank.ai</a></li><li>Twitter: <a href="https://twitter.com/NFTbank_ai">https://twitter.com/NFTbank_ai</a></li><li>Discord: <a href="https://t.co/N2kvmpvko6">discord.gg/8Rg63UnhYb</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ac73e88aa3ad" width="1" height="1" alt=""><hr><p><a href="https://medium.com/nftbank-ai/how-to-accurately-estimate-the-price-of-individual-nfts-part-2-ac73e88aa3ad">How to accurately estimate the price of individual NFTs — Part 2</a> was originally published in <a href="https://medium.com/nftbank-ai">NFTBank.ai</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to value items in NFT projects? — Part 1]]></title>
            <link>https://medium.com/nftbank-ai/how-to-value-items-in-nft-projects-part-1-a87be8bcb21d?source=rss-1b573bb17e68------2</link>
            <guid isPermaLink="false">https://medium.com/p/a87be8bcb21d</guid>
            <dc:creator><![CDATA[Ryker]]></dc:creator>
            <pubDate>Fri, 12 Nov 2021 10:59:38 GMT</pubDate>
            <atom:updated>2024-04-04T16:18:59.680Z</atom:updated>
            <content:encoded><![CDATA[<h3>How to value items in NFT projects? — Part 1</h3><h3>TL’DR: Strong baseline model is how NFTBank provides NFT price estimates with high accuracy</h3><p>After researching various Non-fungible tokens (NFTs) projects, we found that, like <a href="https://medium.com/@Andrew.Steinwold/how-to-value-virtual-assets-d51a0c323709">this article</a>, we cannot apply the same valuation model to all projects, and each project needs a valuation model that suits it.</p><p><strong>However, if we consider the growth of NFTs, the total time required to find perfectly suitable valuation models for the entire NFT market increases exponentially.</strong> Thus, we see the most scalable approach as having a <strong>baseline model</strong> which has a certain level of performance to all NFT projects to simultaneously improve the <strong>coverage</strong> and <strong>depth</strong>.</p><p>This is how we provide estimated price of NFT items in your portfolio, and on the market. For example, the figures below are scatter plots with predicted and actual price for weekly transaction on the test set (data from 2021. 09. 01 to 2021. 10. 03) of the Cryptopunks project. Our baseline model shows a prediction error of about <strong>9.974%</strong> with respect to the actual transaction price in the Cryptopunks project.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wsv4tC4PGxv8pRZ9" /><figcaption>Scatter plot with predicted and actual price (Cryptopunks MAPE: 9.974%)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*oAC3NTVbHXKS01_4" /><figcaption>Scatter plot with predicted and actual price (Cool Cats NFTs MAPE : 10.92%)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bUvHi_KRTevqeVg99E_rxw.png" /><figcaption>Scatter plot with predicted and actual price (Lazy Lions MAPE : 13.65%)</figcaption></figure><p>Even in other projects, including the Cryptopunks project, the baseline model shows a certain level of predictive performance (this will be discussed in more detail). Thus, as soon as the data related to a NFT project is prepared, a baseline model is first fitted and the performance of the model is defined as the baseline (just like its name). Afterwards, we develop and produce a more sophisticated model than baseline to produce an appropriate estimated price with many iterations.</p><h3>Do you want to read more about how we developed our baseline Model? Read on.</h3><h3>Why is NFT price estimation so difficult?</h3><p>As you may all know, extreme price and volume fluctuations are the norm when it comes to NFT projects, or even the entire NFT market.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x7FaBi0oo6dKBKc-xC-oxQ.png" /></figure><p>For example, Bored ape yacht club (BAYC) #3749 was traded for 105 ETH on July 16, and 740 ETH on September 6, 2021, which means that in less than two months, its value rose <strong>705%</strong>. What’s really interesting is that this is not an extreme case. Cool cats #8875 was traded for 0.8 ETH on July 3, and 75 ETH on August 28, 2021, which is a price increase of about <strong>9375%</strong> in less than two months. As such, both overall volume of the NFT projects and price of the each item in projects change more rapidly than those of traditional financial products.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JyREbdOSMrTmmOuhBkqBOQ.png" /></figure><p>While NFT has been one of the hottest topics in 2021, many are lost in how to assess NFTs’ value, making them hesitant to jump on to the lucrative NFT investment opportunities.</p><p>Two main factors make the valuation model an extremely challenging feat:</p><ul><li><strong>low transaction volume</strong></li><li><strong>extreme price change</strong></li></ul><p>Let’s take Cryptopunks, the NFT project considered most prestigious, as an example. Of the 10,000 punks that are out there, only 5878 of them have any transaction record. Since its launch on June 23rd 2017 to October 4th 2021, there are only 18,777 sales records. Even the punks that switched hands two times or more, the average time between sales is 117 days, up to maximum 1549 days. On top of such low transaction volume, the price change is extreme — and we’re not even considering the rare items that have never been available on the market.</p><h3>Why can’t we use the conventional methods ?</h3><p>There is an essential difference between existing financial products and NFTs in terms of liquidity and cash flow. Therefore, it is inappropriate to evaluate the value of NFT products using conventional methods.</p><p>Let’s look at two intuitive valuation methods used for financial products.</p><ul><li>Latest sold price? The market flow at the time the transaction occurred and the current flow may differ significantly. For example, BAYC #9361 traded for 1 ETH in May 1, 2021, but in August 26, it traded for 500 ETH. In other words, it is risky to estimate the “current” value using the latest transaction price.</li><li>Floor price? The price range is significantly different for rare items and non-rare items. As an easy example, the Male type of Cryptopunks is listed on at least 102 ETH. On the other hand, the Alien type is listed on 35,000 ETH. In other words, it is risky to estimate the price of an item itself through the floor price of all current market prices.</li></ul><p>In addition, we found that traditional statistical models, such as time series models, are inadequate for NFTs. Of course, it can be applied, but you need to put a lot of effort into each project to perform hyperparameter tuning to compose a “decent” model to a certain performance level (even worse, you may not be rewarded for your efforts). Also, since NFTs have different characteristics such as collectibles and game items for each project, it will take forever to implement a model suitable for each project considering the growth of NFTs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/577/1*-OcH2n0LhYKEdw_gicOu4g.png" /></figure><p>This is the reason why NFTBank is continuously striving to develop a suitable evaluation model for NFT projects. The ultimate goal of our journey is to produce a model with the best possible performance to all projects to ensure modeling coverage and depth at the same time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lJ97Jm9K_a-93ewJha24UA.png" /></figure><h3>NFTBank’s first steps to baseline model</h3><p>The modeling goal of NFTBank is to construct a general modeling that can be used in various NFTs projects. All NFT data have two essential components. One is on-chain data containing transaction details, and the other is off-chain data containing traits of each item. If there is no previous transaction history, it will be difficult to understand the market flow, and if there is no off-chain data, it will be impossible to grasp the properties of an item such as rarity. NFTbank has the state-of-the-art technology to extract, compose, and serve the two data.</p><p>The two major problems considered in the modeling point of view are as follows.</p><ul><li>How to deal with bundle transaction?</li><li>How to deal with categorical variable and rarity of it?</li></ul><h3>1. Tackling the bundle imputations</h3><p>The bundle transaction is one factor that can affect modeling performance. For example, in the League of kingdoms (LOK) project, let’s assume that several connected lands were traded as bundles (the case of being sold as a bundle among train sets is a very common case exceeding 30%). As in the our previously published LOK-related article (<a href="https://medium.com/nftbank-ai/whats-happening-at-lok-land-market-part-1-understanding-land-rewards-programs-land-markets-d0c054e654d4">part 1</a>, <a href="https://medium.com/nftbank-ai/whats-happening-at-lok-land-market-part-2-find-a-land-token-worth-investing-in-aea37929aca9">part 2</a>, <a href="https://medium.com/nftbank-ai/whats-happening-at-lok-land-market-part-3-best-way-to-purchase-lok-land-98a6fc6ad6f6">part 3</a>), even if it is a model created based on the properties of neighboring lands, if regions with high potential value and regions with low potential are sold in bundles, problems can arise in predicting the value of each land (because, simply, we can’t get the price of each land in bundles).</p><p>We introduce the method of bundle imputation for this problem. Bundle imputation is a method of constructing a model by pre-training the price of land included in the bundles and re-training the data set containing bundle imputed price with another model. The schematic imputation process is shown in the figure follows.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QNeBPIdM5a153FRqQovmjw.png" /></figure><p><a href="https://kroki.io/mermaid/svg/eNplkM1uwjAQhO99in0BDvEDVGpwgBBoOaAKyUKVlSwlan6Qa9QLD1977Aaj3tbfjMee_TT6cqa9fCJ6USS11XSk2eyZcrU3uh2OXgCYq3fdtc0dSHXgb-vPebiRuSsf-XVoOqZDwkXC3zyfh8DMJT74IxcJh1-G91w-P_gjFwmHP88gFGpn2PoW1I8Nd3giSDd00bYdhxsVjhfACxVjyv5ytZCRJ4Lq79_HBaYlWsPPWE7EKJHg-KeVeuUfmla7jNSPYhrjJ0t4p63LQNchIa5-BVap7V_B8n-_yuF1wDs2p9H0eqiZ6jPXX0GtoG78vpq2Ru1fmVGRuw==">Schematic imputation process</a></p><p>As a result of applying bundle imputation techniques, it was possible to obtain a more stable model in terms of standard deviation of MAPE, and it was confirmed that the performance was also improved.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cVtxb5g-8ZHTBKBMuHUnRA.png" /></figure><h3>2. Tackling the categorical features</h3><p>Categorical variables or traits are so diverse in each project that there are often no items with the same combination. Also, if you filter items with the same trait for every categorical variable, the influence of the most important categorical variable in determining the price may decrease. As an easy example, Meebits #10761 has a record of trading for 700 ETH. Although the most important trait possessed by this item is the “type”, there is a problem that the price can be bumped by other general traits possessed if the model learns traits without careful selection.</p><p>This implies that it is difficult to use distance based (e.g. knn) techniques and regression based techniques from a machine learning point of view, and even if used, it is easy to cause problems in scale up because sophisticated weights or feature selection have to be considered. Therefore, we considered a modeling technique based on the gradient boosting model (GBM), which is a representative non-linear model. However, it is difficult to see that the choice of modeling technique is an accurate solution to the two problems above.</p><p>The idea devised by NFTbank as a countermeasure against the first problem is a new embedding feature creation technique. Our embedding feature creation technique started from the intuition of finding a nonlinear transformation that can best explain the selling price with categorical traits. In other words, we estimate the W and f by minimizing the following objective function, and assume that f(x) is an embedding vector that can explain categorical variables well.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/376/1*9OQHlp3KNn7cYlDPLaLqRA.png" /></figure><p>Modeling consists of two stages, first estimating the best nonlinear transformation f that can explain price through traits, transforming traits through estimated f, and evaluating performance by fitting a GBM based model with transaction details. This conversion technique has similar characteristics to the ideas of Autoencoder and Word2vec, and the embedding function f(x) is designed as a sufficiently complex dense neural network.</p><p>If so, how can rarity be reflected? Our model was designed to reflect the rarity that could not be considered in the one-hot encoding method by considering the inverse document frequency weight (idf) in the trait before making the embedding feature. As shown in the table below, it was confirmed that the performance of the proposed model was improved more than that of GBM based on the median absolute percentage error (MAPE) where</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/722/1*IT-cZ86SDVIpsWBVtGrsHA.png" /></figure><p>and this was used in the price evaluation model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*q0cNvRyWSbyYevJFrYR1oQ.png" /></figure><p>In the table above, you can easily understand the difficulties of constructing model through the Bonsai case. When penalized linear models such as Ridge and Lasso, which are representative statistical linear models, are used, the error rates are 70% and 48%, so no one would believe a predictor with such an error.</p><h3>So is this how NFTBank’s baseline model works now?</h3><p>Sorry to those who have read this far, but the introduced techniques above are not the latest baseline model (but included in). The proposed model has been continuously improved and will be improved while solving problems that occur as the number of covered projects increases. The generalized optimization problem of the model at this time is as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/562/1*6Y1Qr8NKj8mR8ZlY43_kQQ.png" /></figure><p>Although the form of the specific function cannot be revealed 🥲, f is a bijective function and g is a GBM-based prediction model. After the tree model is trained, the predicted value is provided to the product through the closed form of below formula since f has inverse function always by definition.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/354/1*w1wDD8wN_FfAcTcipayWAw.png" /></figure><p>The performance of some projects with baseline model is as follows.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U4Kp-zz-oJ7QqHBZtIfG7Q.png" /></figure><p>Below is the scatter plot with predicted and actual price of the Veefriends and Meebits projects. Of course, there will be some error, but it can be seen that the difference between the predicted value and the actual value is small in most data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n_21LYAXfGYn8MLyNYofpg.png" /><figcaption>Scatter plot on Veefriends project.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cNS9J38f6qAXBWar3WRAEw.png" /><figcaption>Scatter plot on Meebits project.</figcaption></figure><h3>Limitation</h3><p>Even if our model is an improved form from a machine learning point of view, we have experienced various cases where our predictions deviate significantly, and we are always thinking about a breakthrough. Some limitations of the current model and our improvement goals are as follows.</p><p>Our model, as most machine learning models, is a model that estimates conditional expectation. Quite simply, the “average” value of items that share the same trait on the market today can be significantly <strong>under</strong> or <strong>over</strong> estimated from the value thought by the participants of the NFT project 😭. For example, in gaming NFT, it will be difficult to predict the change of the market flow with the average market value formed through numerous previous transactions in a situation where an undervalued item suddenly rises as the meta changes.</p><p>As another example, our model is difficult to respond to game changers. We define a game changer as a rare item suddenly traded in the market. For example, in the Cryptopunks project, if an alien typed item was first traded for 4200 ETH, it is impossible for our model to return an adequate forecast for it, as there is no history of previous sales.</p><p>Our model does not take the floor price into account. The floor price is defined by the minimum price among the currently “listed” prices. We understand that floor price is a very important indicator in the NFT scene. But, when considering the floor price, there is a high possibility that the estimation will be heavily dumped (either increasing or decreasing) since it is not “traded” prices. Therefore, there is a possibility that the predicted value of the model may be lower than the floor price 😰.</p><p>As of now, the baseline model is not applied to all projects. This is because, as you know, the data required for modeling is not magically created. It takes time and effort to prepare data, so it is conquered one by one. We want you to know that we are putting a lot of effort and time into it.</p><p>Finally, the following functions have been added to receive feedback about estimated price for improvement.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IJBRlRXf21yPeTK1yPxTkA.png" /></figure><p>If you have any questions and doubts about the estimated price, please feel free to give us feedback. Each and every feedback is checked, and they will definitely be reflected for further improved modeling.</p><h3>Conclusion</h3><p>The model performance of NFTBank is constantly being improved through numerous experiments and efforts, and it will not stop until it can be applied to all projects. There are countless theorems in the field of optimization and machine learning, but our favorite is “No free lunch theorem”. This theorem implies that our model may not perform best for all projects. It also implies that the next step for NFTBank is to develop and produce a model that better fits each project after mastering our baseline model.</p><p>In a slightly different sense, there is no free lunch and even no free dinner at NFTBank. All team members are working day and night to make NFTBank the most effective tool for managing NFT assets. At last, a big thank to our users for supporting us and giving us great feedback. This is a huge motivator for us and means a lot to us. Stay tuned!</p><h3>Join our socials</h3><ul><li>Official website: <a href="https://nftbank.ai/">https://nftbank.ai</a></li><li>Twitter: <a href="https://twitter.com/NFTbank_ai">https://twitter.com/NFTbank_ai</a></li><li>Discord: <a href="https://t.co/N2kvmpvko6">discord.gg/8Rg63UnhYb</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a87be8bcb21d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/nftbank-ai/how-to-value-items-in-nft-projects-part-1-a87be8bcb21d">How to value items in NFT projects? — Part 1</a> was originally published in <a href="https://medium.com/nftbank-ai">NFTBank.ai</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>